| POST | /v1/chat/completions |
|---|
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Messages | body | List<AiMessage> | Yes | The messages to generate chat completions for. |
| Model | body | string | Yes | ID of the model to use. See the model endpoint compatibility table for details on which models work with the Chat API |
| Audio | body | AiChatAudio | No | Parameters for audio output. Required when audio output is requested with modalities: [audio] |
| LogitBias | body | Dictionary<int, int> | No | Modify the likelihood of specified tokens appearing in the completion. |
| Metadata | body | Dictionary<string, string> | No | Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format. |
| ReasoningEffort | body | string | No | Constrains effort on reasoning for reasoning models. Currently supported values are minimal, low, medium, and high (none, default). Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.Allowable Values
|
| ResponseFormat | body | AiResponseFormat | No | An object specifying the format that the model must output. Compatible with GPT-4 Turbo and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`. Setting Type to ResponseFormat.JsonObject enables JSON mode, which guarantees the message the model generates is valid JSON. |
| ServiceTier | body | string | No | Specifies the processing type used for serving the request.Allowable Values
|
| SafetyIdentifier | body | string | No | A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user. |
| Stop | body | List<string> | No | Up to 4 sequences where the API will stop generating further tokens. |
| Modalities | body | List<string> | No | Output types that you would like the model to generate. Most models are capable of generating text, which is the default: |
| PromptCacheKey | body | string | No | Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. |
| Tools | body | List<Tool> | No | A list of tools the model may call. Currently, only functions are supported as a tool. Use this to provide a list of functions the model may generate JSON inputs for. A max of 128 functions are supported. |
| Verbosity | body | string | No | Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high.Allowable Values
|
| Temperature | body | double? | No | What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.Valid Range: 0 - 2 |
| MaxCompletionTokens | body | int? | No | An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.Valid Range: 1 - |
| TopLogprobs | body | int? | No | An integer between 0 and 20 specifying the number of most likely tokens to return at each token position, each with an associated log probability. logprobs must be set to true if this parameter is used.Valid Range: 0 - 20 |
| TopP | body | double? | No | An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.Valid Range: 0 - 1 |
| FrequencyPenalty | body | double? | No | Number between `-2.0` and `2.0`. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.Valid Range: 0 - 2 |
| PresencePenalty | body | double? | No | Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.Valid Range: 0 - 2 |
| Seed | body | int? | No | This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result. Determinism is not guaranteed, and you should refer to the system_fingerprint response parameter to monitor changes in the backend. |
| N | body | int? | No | How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep `n` as `1` to minimize costs. |
| Store | body | bool? | No | Whether or not to store the output of this chat completion request for use in our model distillation or evals products. |
| Logprobs | body | bool? | No | Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message. |
| ParallelToolCalls | body | bool? | No | Whether to enable parallel function calling during tool use. |
| EnableThinking | body | bool? | No | Whether to enable thinking mode for some Qwen models and providers. |
| Stream | body | bool? | No | If set, partial message deltas will be sent, like in ChatGPT. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a `data: [DONE]` message. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Content | form | List<AiContent> | No | The contents of the message. |
| Role | form | string | Yes | The role of the author of this message. Valid values are `system`, `user`, `assistant` and `tool`. |
| Name | form | string | No | An optional name for the participant. Provides the model information to differentiate between participants of the same role. |
| ToolCalls | form | List<ToolCall> | No | The tool calls generated by the model, such as function calls. |
| ToolCallId | form | string | No | Tool call that this message is responding to. |
| Reasoning | form | string | No | The reasoning an assistant message was generated with, normalized per provider when replayed as history. |
| ReasoningContent | form | string | No | The reasoning an assistant message was generated with, as emitted by Gemini and most OpenAI-compatible providers. |
| Timestamp | form | long? | No | Unix timestamp (in milliseconds) the message was generated. |
| Images | form | List<AiContent> | No | Images attached to the message. Folded into `content` parts before sending to a provider. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Type | form | string | Yes | The type of the content part. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Id | form | string | Yes | The ID of the tool call. |
| Type | form | string | Yes | The type of the tool. Currently, only `function` is supported. |
| Function | form | ToolFunction | Yes | The function that the model called. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Name | form | string | Yes | The name of the function to call. |
| Arguments | form | string | Yes | The arguments to call the function with, as generated by the model in JSON format. Note that the model does not always generate valid JSON, and may hallucinate parameters not defined by your function schema. Validate the arguments in your code before calling your function. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Format | form | string | Yes | Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16. |
| Voice | form | string | Yes | The voice the model uses to respond. Supported voices are alloy, ash, ballad, coral, echo, fable, nova, onyx, sage, and shimmer. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Type | form | ResponseFormat | No | An object specifying the format that the model must output. Compatible with GPT-4 Turbo and all GPT-3.5 Turbo models newer than gpt-3.5-turbo-1106. |
| Name | Value | |
|---|---|---|
| Text | text | |
| JsonObject | json_object |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Type | form | ToolType | No | The type of the tool. Currently, only function is supported. |
| Function | form | AiToolFunction | No | The function definition the model may call. |
| Name | Value | |
|---|---|---|
| Function | function |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Name | form | string | No | The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64. |
| Description | form | string | No | A description of what the function does, used by the model to choose when and how to call the function. |
| Parameters | form | Dictionary<string, Object> | No | The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Id | form | string | Yes | A unique identifier for the chat completion. |
| Choices | form | List<Choice> | Yes | A list of chat completion choices. Can be more than one if n is greater than 1. |
| Created | form | long | No | The Unix timestamp (in seconds) of when the chat completion was created. |
| Model | form | string | Yes | The model used for the chat completion. |
| SystemFingerprint | form | string | No | This fingerprint represents the backend configuration that the model runs with. |
| Object | form | string | Yes | The object type, which is always chat.completion. |
| ServiceTier | form | string | No | Specifies the processing type used for serving the request. |
| Usage | form | AiUsage | Yes | Usage statistics for the completion request. |
| Provider | form | string | No | The provider used for the chat completion. |
| Cost | form | double? | No | Total cost of the completion in USD, accumulated across every request in the tool loop. |
| ToolHistory | form | List<ChoiceMessage> | No | The assistant and tool messages exchanged during the tool-execution loop, in order. |
| Metadata | form | Dictionary<string, string> | No | Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format. |
| ResponseStatus | form | ResponseStatus | No |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| FinishReason | form | string | Yes | The reason the model stopped generating tokens. This will be stop if the model hit a natural stop point or a provided stop sequence, length if the maximum number of tokens specified in the request was reached, content_filter if content was omitted due to a flag from our content filters, tool_calls if the model called a tool |
| Index | form | int | No | The index of the choice in the list of choices. |
| Message | form | ChoiceMessage | Yes | A chat completion message generated by the model. |
| Logprobs | form | Logprobs | No | Log probability information for the choice. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Content | form | string | Yes | The contents of the message. |
| Refusal | form | string | No | The refusal message generated by the model. |
| Reasoning | form | string | No | The reasoning process used by the model. |
| ReasoningContent | form | string | No | The reasoning process used by the model, as emitted by Gemini and most OpenAI-compatible providers. |
| Thinking | form | string | No | The reasoning process used by the model, as emitted by Anthropic. |
| Role | form | string | Yes | The role of the author of this message. |
| Timestamp | form | long? | No | Unix timestamp (in milliseconds) the message was generated. |
| ToolCallId | form | string | No | The tool call this message is responding to, set on `tool` role messages in tool_history. |
| Images | form | List<AiContent> | No | Images generated by the model or produced by a tool call. |
| Audios | form | List<AiContent> | No | Audio generated by the model or produced by a tool call. |
| Files | form | List<AiContent> | No | Files produced by a tool call. |
| Annotations | form | List<ChoiceAnnotation> | No | Annotations for the message, when applicable, as when using the web search tool. |
| Audio | form | ChoiceAudio | No | If the audio output modality is requested, this object contains data about the audio response from the model. |
| ToolCalls | form | List<ToolCall> | No | The tool calls generated by the model, such as function calls. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Type | form | string | Yes | The type of the URL citation. Always url_citation. |
| UrlCitation | form | UrlCitation | Yes | A URL citation when using web search. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| EndIndex | form | int | No | The index of the last character of the URL citation in the message. |
| StartIndex | form | int | No | The index of the first character of the URL citation in the message. |
| Title | form | string | Yes | The title of the web resource. |
| Url | form | string | Yes | The URL of the web resource. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Data | form | string | Yes | Base64 encoded audio bytes generated by the model, in the format specified in the request. |
| ExpiresAt | form | long | No | The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations. |
| Id | form | string | Yes | Unique identifier for this audio response. |
| Transcript | form | string | Yes | Transcript of the audio generated by the model. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Content | form | List<LogprobItem> | Yes | A list of message content tokens with log probability information. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| Token | form | string | Yes | The token. |
| Logprob | form | double | No | The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value `-9999`.0 is used to signify that the token is very unlikely. |
| Bytes | form | byte[] | Yes | A list of integers representing the UTF-8 bytes representation of the token. Useful in instances where characters are represented by multiple tokens and their byte representations must be combined to generate the correct text representation. Can be `null` if there is no bytes representation for the token. |
| TopLogprobs | form | List<LogprobItem> | Yes | List of the most likely tokens and their log probability, at this token position. In rare cases, there may be fewer than the number of requested `top_logprobs` returned. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| CompletionTokens | form | long | No | Number of tokens in the generated completion. |
| PromptTokens | form | long | No | Number of tokens in the prompt. |
| TotalTokens | form | long | No | Total number of tokens used in the request (prompt + completion). |
| CompletionTokensDetails | form | AiCompletionUsage | No | Breakdown of tokens used in a completion. |
| PromptTokensDetails | form | AiPromptUsage | No | Breakdown of tokens used in the prompt. |
| Duration | form | long? | No | Seconds spent servicing the completion, including every request in the tool loop. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| AcceptedPredictionTokens | form | long | No | When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion. |
| AudioTokens | form | long | No | Audio input tokens generated by the model. |
| ReasoningTokens | form | long | No | Tokens generated by the model for reasoning. |
| RejectedPredictionTokens | form | long | No | When using Predicted Outputs, the number of tokens in the prediction that did not appear in the completion. |
| Name | Parameter | Data Type | Required | Description |
|---|---|---|---|---|
| AcceptedPredictionTokens | form | long | No | When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion. |
| AudioTokens | form | long | No | Audio input tokens present in the prompt. |
| CachedTokens | form | long | No | Cached tokens present in the prompt. |
To override the Content-type in your clients, use the HTTP Accept Header, append the .json suffix or ?format=json
To embed the response in a jsonp callback, append ?callback=myCallback
The following are sample HTTP requests and responses. The placeholders shown need to be replaced with actual values.
POST /v1/chat/completions HTTP/1.1
Host: blazor-vue.web-templates.io
Accept: application/json
Content-Type: application/json
Content-Length: length
{"messages":[{"content":[{}],"role":"String","name":"String","tool_calls":[{"id":"String","type":"String","function":{"name":"String","arguments":"String"}}],"tool_call_id":"String","reasoning":"String","reasoning_content":"String","timestamp":0,"images":[{}]}],"model":"String","audio":{"format":"String","voice":"String"},"logit_bias":{"0":0},"metadata":{"String":"String"},"reasoning_effort":"String","response_format":{"type":"text"},"service_tier":"String","safety_identifier":"String","stop":["String"],"modalities":["String"],"prompt_cache_key":"String","tools":[{"type":"function","function":{"name":"String","description":"String","parameters":{"String":{}}}}],"verbosity":"String","temperature":0,"max_completion_tokens":0,"top_logprobs":0,"top_p":0,"frequency_penalty":0,"presence_penalty":0,"seed":0,"n":0,"store":false,"logprobs":false,"parallel_tool_calls":false,"enable_thinking":false,"stream":false}
HTTP/1.1 200 OK
Content-Type: application/json
Content-Length: length
{"id":"String","choices":[{"finish_reason":"String","index":0,"message":{"content":"String","refusal":"String","reasoning":"String","reasoning_content":"String","thinking":"String","role":"String","timestamp":0,"tool_call_id":"String","images":[{}],"audios":[{}],"files":[{}],"annotations":[{"type":"String","url_citation":{"end_index":0,"start_index":0,"title":"String","url":"String"}}],"audio":{"data":"String","expires_at":0,"id":"String","transcript":"String"},"tool_calls":[{"id":"String","type":"String","function":{"name":"String","arguments":"String"}}]},"logprobs":{"content":[{"token":"String","logprob":0,"bytes":"AA==","top_logprobs":[{"token":"String","logprob":0,"bytes":"AA==","top_logprobs":[{"token":"String","logprob":0,"bytes":"AA=="}]}]}]}}],"created":0,"model":"String","system_fingerprint":"String","object":"String","service_tier":"String","usage":{"completion_tokens":0,"prompt_tokens":0,"total_tokens":0,"completion_tokens_details":{"accepted_prediction_tokens":0,"audio_tokens":0,"reasoning_tokens":0,"rejected_prediction_tokens":0},"prompt_tokens_details":{"accepted_prediction_tokens":0,"audio_tokens":0,"cached_tokens":0},"duration":0},"provider":"String","cost":0,"tool_history":[{"content":"String","refusal":"String","reasoning":"String","reasoning_content":"String","thinking":"String","role":"String","timestamp":0,"tool_call_id":"String","images":[{}],"audios":[{}],"files":[{}],"annotations":[{"type":"String","url_citation":{"end_index":0,"start_index":0,"title":"String","url":"String"}}],"audio":{"data":"String","expires_at":0,"id":"String","transcript":"String"},"tool_calls":[{"id":"String","type":"String","function":{"name":"String","arguments":"String"}}]}],"metadata":{"String":"String"},"responseStatus":{"errorCode":"String","message":"String","stackTrace":"String","errors":[{"errorCode":"String","fieldName":"String","message":"String","meta":{"String":"String"}}],"meta":{"String":"String"}}}