Azure OpenAI in Microsoft Foundry Models v1 REST API reference

Learn how to use Azure OpenAI’s v1 REST API.

Azure OpenAI in Microsoft Foundry Models v1 REST API reference

Only a subset of operations are currently supported with the v1 API. To learn more, see the API version lifecycle guide.

• v1 OpenAPI 3.0 spec

Create chat completion

POST {endpoint}/openai/v1/chat/completions

Creates a chat completion.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Microsoft Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
audio	object	Parameters for audio output. Required when audio output is requested with modalities: ["audio"].	No
└─ format	enum	Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16. Possible values: wav, aac, mp3, flac, opus, pcm16	No
└─ voice	object		No
data_sources	array	The data sources to use for the On Your Data feature, exclusive to Azure OpenAI.	No
frequency_penalty	number	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.	No	0
function_call	enum	Specifying a particular function via {"name": "my_function"} forces the model to call that function. Possible values: none, auto	No
functions	array	Deprecated in favor of tools. A list of functions the model may generate JSON inputs for.	No
logit_bias	object	Modify the likelihood of specified tokens appearing in the completion. Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.	No	None
logprobs	boolean	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.	No	False
max_completion_tokens	integer	An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.	No
max_tokens	integer	The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API. This value is now deprecated in favor of max_completion_tokens, and is not compatible with o1 series models.	No
messages	array	A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text, images, and audio.	Yes
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
modalities	object	Output types that you would like the model to generate. Most models are capable of generating text, which is the default: ["text"] The gpt-4o-audio-preview model can also be used to generate audio. To request that this model generate both text and audio responses, you can use: ["text", "audio"]	No
model	string	The model deployment identifier to use for the chat completion request.	Yes
n	integer	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.	No	1
parallel_tool_calls	object	Whether to enable parallel function calling during tool use.	No
prediction	object	Base representation of predicted output from a model.	No
└─ type	OpenAI.ChatOutputPredictionType		No
presence_penalty	number	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.	No	0
reasoning_effort	object	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
response_format	object		No
└─ type	enum	Possible values: text, json_object, json_schema	No
seed	integer	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.	No
stop	object	Not supported with latest reasoning models o3 and o4-mini. Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.	No
store	boolean	Whether or not to store the output of this chat completion request for use in model distillation or evals products.	No	False
stream	boolean	If set to true, the model response data will be streamed to the client as it is generated using server-sent events.	No	False
stream_options	object	Options for streaming response. Only set this when you set stream: true.	No
└─ include_usage	boolean	If set, an additional chunk will be streamed before the data: [DONE] message. The usage field on this chunk shows the token usage statistics for the entire request, and the choices field will always be an empty array. All other chunks will also include a usage field, but with a null value. NOTE: If the stream is interrupted, you may not receive the final usage chunk which contains the total token usage for the request.	No
temperature	number	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. We generally recommend altering this or top_p but not both.	No	1
tool_choice	OpenAI.ChatCompletionToolChoiceOption	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool. none is the default when no tools are present. auto is the default if tools are present.	No
tools	array	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.	No
top_logprobs	integer	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.	No
top_p	number	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. We generally recommend altering this or temperature but not both.	No	1
user	string	A unique identifier representing your end-user, which can help to monitor and detect abuse.	No
user_security_context	AzureUserSecurityContext	User security context contains several parameters that describe the application itself, and the end user that interacts with the application. These fields assist your security operations teams to investigate and mitigate security incidents by providing a comprehensive approach to protecting your AI applications. Learn more about protecting AI applications using Microsoft Defender for Cloud.	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	AzureCreateChatCompletionResponse
text/event-stream	AzureCreateChatCompletionStreamResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Examples

Example

Creates a completion for the provided prompt, parameters and chosen model.

POST {endpoint}/openai/v1/chat/completions

{
 "model": "gpt-4o-mini",
 "messages": [
  {
   "role": "system",
   "content": "you are a helpful assistant that talks like a pirate"
  },
  {
   "role": "user",
   "content": "can you tell me how to care for a parrot?"
  }
 ]
}

Responses: Status Code: 200

{
  "body": {
    "id": "chatcmpl-7R1nGnsXO8n4oi9UPz2f3UHdgAYMn",
    "created": 1686676106,
    "choices": [
      {
        "index": 0,
        "finish_reason": "stop",
        "message": {
          "role": "assistant",
          "content": "Ahoy matey! So ye be wantin' to care for a fine squawkin' parrot, eh?..."
        }
      }
    ],
    "usage": {
      "completion_tokens": 557,
      "prompt_tokens": 33,
      "total_tokens": 590
    }
  }
}

Create embedding

POST {endpoint}/openai/v1/embeddings

Creates an embedding vector representing the input text.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
dimensions	integer	The number of dimensions the resulting output embeddings should have. Only supported in text-embedding-3 and later models.	No
encoding_format	enum	The format to return the embeddings in. Can be either float or base64. Possible values: float, base64	No
input	string or array		Yes
model	string	The model to use for the embedding request.	Yes
user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.CreateEmbeddingResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Examples

Example

Return the embeddings for a given prompt.

POST {endpoint}/openai/v1/embeddings

{
 "model": "text-embedding-ada-002",
 "input": [
  "this is a test"
 ]
}

Responses: Status Code: 200

{
  "body": {
    "data": [
      {
        "index": 0,
        "embedding": [
          -0.012838088,
          -0.007421397,
          -0.017617522,
          -0.028278312,
          -0.018666342,
          0.01737855,
          -0.01821495,
          -0.006950092,
          -0.009937238,
          -0.038580645,
          0.010674067,
          0.02412286,
          -0.013647936,
          0.013189907,
          0.0021125758,
          0.012406612,
          0.020790534,
          0.00074595667,
          0.008397198,
          -0.00535031,
          0.008968075,
          0.014351576,
          -0.014086051,
          0.015055214,
          -0.022211088,
          -0.025198232,
          0.0065186154,
          -0.036350243,
          0.009180495,
          -0.009698266,
          0.009446018,
          -0.008463579,
          -0.0040426035,
          -0.03443847,
          -0.00091273896,
          -0.0019217303,
          0.002349888,
          -0.021560553,
          0.016515596,
          -0.015572986,
          0.0038666942,
          -8.432463e-05,
          0.0032178196,
          -0.020365695,
          -0.009631885,
          -0.007647093,
          0.0033837722,
          -0.026764825,
          -0.010501476,
          0.020219658,
          0.024640633,
          -0.0066912062,
          -0.036456455,
          -0.0040923897,
          -0.013966565,
          0.017816665,
          0.005366905,
          0.022835068,
          0.0103488,
          -0.0010811808,
          -0.028942121,
          0.0074280356,
          -0.017033368,
          0.0074877786,
          0.021640211,
          0.002499245,
          0.013316032,
          0.0021524043,
          0.010129742,
          0.0054731146,
          0.03143805,
          0.014856071,
          0.0023366117,
          -0.0008243692,
          0.022781964,
          0.003038591,
          -0.017617522,
          0.0013309394,
          0.0022154662,
          0.00097414135,
          0.012041516,
          -0.027906578,
          -0.023817508,
          0.013302756,
          -0.003003741,
          -0.006890349,
          0.0016744611
        ]
      }
    ],
    "usage": {
      "prompt_tokens": 4,
      "total_tokens": 4
    }
  }
}

List evals

GET {endpoint}/openai/v1/evals

List evaluations for a project.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
after	query	No	string	Identifier for the last eval from the previous pagination request.
limit	query	No	integer	A limit on the number of evals to be returned in a single pagination response.
order	query	No	string Possible values: asc, desc	Sort order for evals by timestamp. Use asc for ascending order or desc for descending order.
order_by	query	No	string Possible values: created_at, updated_at	Evals can be ordered by creation time or last updated time. Use created_at for creation time or updated_at for last updated time.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.EvalList

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create eval

POST {endpoint}/openai/v1/evals

Create the structure of an evaluation that can be used to test a model’s performance.

An evaluation is a set of testing criteria and a datasource. After creating an evaluation, you can run it on different models and model parameters. We support several types of graders and datasources.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
data_source_config	object		Yes
└─ type	OpenAI.EvalDataSourceConfigType		No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the evaluation.	No
statusCode	enum	Possible values: 201	Yes
testing_criteria	array	A list of graders for all eval runs in this group. Graders can reference variables in the data source using double curly braces notation, like {{item.variable_name}}. To reference the model’s output, use the sample namespace (ie, {{sample.output_text}}).	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.Eval

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get eval

GET {endpoint}/openai/v1/evals/{eval_id}

Retrieve an evaluation by its ID. Retrieves an evaluation by its ID.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.Eval

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Update eval

POST {endpoint}/openai/v1/evals/{eval_id}

Update select, mutable properties of a specified evaluation.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
metadata	OpenAI.MetadataPropertyForRequest	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string		No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.Eval

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Delete eval

DELETE {endpoint}/openai/v1/evals/{eval_id}

Delete a specified evaluation.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	object

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get eval runs

GET {endpoint}/openai/v1/evals/{eval_id}/runs

Retrieve a list of runs for a specified evaluation.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string
after	query	No	string
limit	query	No	integer
order	query	No	string Possible values: asc, desc
status	query	No	string Possible values: queued, in_progress, completed, canceled, failed

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.EvalRunList

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create eval run

POST {endpoint}/openai/v1/evals/{eval_id}/runs

Create a new evaluation run, beginning the grading process.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
data_source	object		Yes
└─ type	OpenAI.EvalRunDataSourceType		No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the run.	No

Responses

Status Code: 201

Description: The request has succeeded and a new resource has been created as a result.

Content-Type	Type	Description
application/json	OpenAI.EvalRun

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get eval run

GET {endpoint}/openai/v1/evals/{eval_id}/runs/{run_id}

Retrieve a specific evaluation run by its ID.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string
run_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.EvalRun

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Cancel eval run

POST {endpoint}/openai/v1/evals/{eval_id}/runs/{run_id}

Cancel a specific evaluation run by its ID.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string
run_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.EvalRun

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Delete eval run

DELETE {endpoint}/openai/v1/evals/{eval_id}/runs/{run_id}

Delete a specific evaluation run by its ID.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string
run_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	object

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get eval run output items

GET {endpoint}/openai/v1/evals/{eval_id}/runs/{run_id}/output_items

Get a list of output items for a specified evaluation run.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string
run_id	path	Yes	string
after	query	No	string
limit	query	No	integer
status	query	No	string Possible values: fail, pass
order	query	No	string Possible values: asc, desc

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.EvalRunOutputItemList

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get eval run output item

GET {endpoint}/openai/v1/evals/{eval_id}/runs/{run_id}/output_items/{output_item_id}

Retrieve a specific output item from an evaluation run by its ID.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-evals	header	Yes	string Possible values: preview	Enables access to AOAI Evals, a preview feature. This feature requires the ‘aoai-evals’ header to be set to ‘preview’.
eval_id	path	Yes	string
run_id	path	Yes	string
output_item_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.EvalRunOutputItem

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create file

POST {endpoint}/openai/v1/files

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: multipart/form-data

Name	Type	Description	Required	Default
expires_after	object		Yes
└─ anchor	AzureFileExpiryAnchor		No
└─ seconds	integer		No
file	string		Yes
purpose	enum	The intended purpose of the uploaded file. One of: - assistants: Used in the Assistants API - batch: Used in the Batch API - fine-tune: Used for fine-tuning - evals: Used for eval data sets Possible values: assistants, batch, fine-tune, evals	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	AzureOpenAIFile

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Examples

Example

POST {endpoint}/openai/v1/files

List files

GET {endpoint}/openai/v1/files

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
purpose	query	No	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	AzureListFilesResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Retrieve file

GET {endpoint}/openai/v1/files/{file_id}

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
file_id	path	Yes	string	The ID of the file to use for this request.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	AzureOpenAIFile

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Delete file

DELETE {endpoint}/openai/v1/files/{file_id}

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
file_id	path	Yes	string	The ID of the file to use for this request.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.DeleteFileResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Download file

GET {endpoint}/openai/v1/files/{file_id}/content

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
file_id	path	Yes	string	The ID of the file to use for this request.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/octet-stream	string

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Run grader

POST {endpoint}/openai/v1/fine_tuning/alpha/graders/run

Run a grader.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
grader	object	A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.	Yes
└─ calculate_output	string	A formula to calculate the output based on grader results.	No
└─ evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	No
└─ graders	object		No
└─ image_tag	string	The image tag to use for the python script.	No
└─ input	array	The input text. This may include template strings.	No
└─ model	string	The model to use for the evaluation.	No
└─ name	string	The name of the grader.	No
└─ operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	No
└─ range	array	The range of the score. Defaults to [0, 1].	No
└─ reference	string	The text being graded against.	No
└─ sampling_params		The sampling parameters for the model.	No
└─ source	string	The source code of the python script.	No
└─ type	enum	The object type, which is always multi. Possible values: multi	No
item		The dataset item provided to the grader. This will be used to populate the item namespace. See the guide for more details.	No
model_sample	string	The model sample to be evaluated. This value will be used to populate the sample namespace. See the guide for more details. The output_json variable will be populated if the model sample is a valid JSON string.	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.RunGraderResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Validate grader

POST {endpoint}/openai/v1/fine_tuning/alpha/graders/validate

Validate a grader.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
grader	object	A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.	Yes
└─ calculate_output	string	A formula to calculate the output based on grader results.	No
└─ evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	No
└─ graders	object		No
└─ image_tag	string	The image tag to use for the python script.	No
└─ input	array	The input text. This may include template strings.	No
└─ model	string	The model to use for the evaluation.	No
└─ name	string	The name of the grader.	No
└─ operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	No
└─ range	array	The range of the score. Defaults to [0, 1].	No
└─ reference	string	The text being graded against.	No
└─ sampling_params		The sampling parameters for the model.	No
└─ source	string	The source code of the python script.	No
└─ type	enum	The object type, which is always multi. Possible values: multi	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ValidateGraderResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create fine-tuning job

POST {endpoint}/openai/v1/fine_tuning/jobs

Creates a fine-tuning job which begins the process of creating a new model from a given dataset.

Response includes details of the enqueued job including job status and the name of the fine-tuned models once complete.

Learn more about fine-tuning

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
hyperparameters	object	The hyperparameters used for the fine-tuning job. This value is now deprecated in favor of method, and should be passed in under the method parameter.	No
└─ batch_size	enum	Possible values: auto	No
└─ learning_rate_multiplier	enum	Possible values: auto	No
└─ n_epochs	enum	Possible values: auto	No
integrations	array	A list of integrations to enable for your fine-tuning job.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
method	OpenAI.FineTuneMethod	The method used for fine-tuning.	No
model	string (see valid models below)	The name of the model to fine-tune. You can select one of the supported models.	Yes
seed	integer	The seed controls the reproducibility of the job. Passing in the same seed and job parameters should produce the same results, but may differ in rare cases. If a seed is not specified, one will be generated for you.	No
suffix	string	A string of up to 64 characters that will be added to your fine-tuned model name. For example, a suffix of “custom-model-name” would produce a model name like ft:gpt-4o-mini:openai:custom-model-name:7p4lURel.	No	None
training_file	string	The ID of an uploaded file that contains training data. See upload file for how to upload a file. Your dataset must be formatted as a JSONL file. Additionally, you must upload your file with the purpose fine-tune. The contents of the file should differ depending on if the model uses the chat, or if the fine-tuning method uses the preference format. See the fine-tuning guide for more details.	Yes
validation_file	string	The ID of an uploaded file that contains validation data. If you provide this file, the data is used to generate validation metrics periodically during fine-tuning. These metrics can be viewed in the fine-tuning results file. The same data should not be present in both train and validation files. Your dataset must be formatted as a JSONL file. You must upload your file with the purpose fine-tune. See the fine-tuning guide for more details.	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.FineTuningJob

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List paginated fine-tuning jobs

GET {endpoint}/openai/v1/fine_tuning/jobs

List your organization’s fine-tuning jobs

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
after	query	No	string	Identifier for the last job from the previous pagination request.
limit	query	No	integer	Number of fine-tuning jobs to retrieve.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListPaginatedFineTuningJobsResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Retrieve fine-tuning job

GET {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}

Get info about a fine-tuning job.

Learn more about fine-tuning

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
fine_tuning_job_id	path	Yes	string	The ID of the fine-tuning job.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.FineTuningJob

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Cancel fine-tuning job

POST {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/cancel

Immediately cancel a fine-tune job.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
fine_tuning_job_id	path	Yes	string	The ID of the fine-tuning job to cancel.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.FineTuningJob

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List fine-tuning job checkpoints

GET {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/checkpoints

List the checkpoints for a fine-tuning job.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
fine_tuning_job_id	path	Yes	string	The ID of the fine-tuning job to get checkpoints for.
after	query	No	string	Identifier for the last checkpoint ID from the previous pagination request.
limit	query	No	integer	Number of checkpoints to retrieve.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListFineTuningJobCheckpointsResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Fine-tuning - Copy checkpoint

POST {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/checkpoints/{fine_tuning_checkpoint_name}/copy

Creates a copy of a fine-tuning checkpoint at the given destination account and region.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-copy-ft-checkpoints	header	Yes	string Possible values: preview	Enables access to checkpoint copy operations for models, an AOAI preview feature. This feature requires the ‘aoai-copy-ft-checkpoints’ header to be set to ‘preview’.
accept	header	Yes	string Possible values: application/json
fine_tuning_job_id	path	Yes	string
fine_tuning_checkpoint_name	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
destinationResourceId	string	The ID of the destination Resource to copy.	Yes
region	string	The region to copy the model to.	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	CopyModelResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Fine-tuning - Get checkpoint

GET {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/checkpoints/{fine_tuning_checkpoint_name}/copy

Gets the status of a fine-tuning checkpoint copy.

This Azure OpenAI operation is in preview and subject to change.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
aoai-copy-ft-checkpoints	header	Yes	string Possible values: preview	Enables access to checkpoint copy operations for models, an AOAI preview feature. This feature requires the ‘aoai-copy-ft-checkpoints’ header to be set to ‘preview’.
accept	header	Yes	string Possible values: application/json
fine_tuning_job_id	path	Yes	string
fine_tuning_checkpoint_name	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	CopyModelResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List fine-tuning events

GET {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/events

Get status updates for a fine-tuning job.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
fine_tuning_job_id	path	Yes	string	The ID of the fine-tuning job to get events for.
after	query	No	string	Identifier for the last event from the previous pagination request.
limit	query	No	integer	Number of events to retrieve.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListFineTuningJobEventsResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Pause fine-tuning job

POST {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/pause

Pause a fine-tune job.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
fine_tuning_job_id	path	Yes	string	The ID of the fine-tuning job to pause.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.FineTuningJob

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Resume fine-tuning job

POST {endpoint}/openai/v1/fine_tuning/jobs/{fine_tuning_job_id}/resume

Resume a paused fine-tune job.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
fine_tuning_job_id	path	Yes	string	The ID of the fine-tuning job to resume.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.FineTuningJob

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List models

GET {endpoint}/openai/v1/models

Lists the currently available models, and provides basic information about each one such as the owner and availability.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListModelsResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Retrieve model

GET {endpoint}/openai/v1/models/{model}

Retrieves a model instance, providing basic information about the model such as the owner and permissioning.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
model	path	Yes	string	The ID of the model to use for this request.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.Model

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create response

POST {endpoint}/openai/v1/responses

Creates a model response.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
background	boolean	Whether to run the model response in the background. Learn more.	No	False
include	array	Specify additional output data to include in the model response. Currently supported values are: - code_interpreter_call.outputs: Includes the outputs of python code execution in code interpreter tool call items. - computer_call_output.output.image_url: Include image urls from the computer call output. - file_search_call.results: Include the search results of the file search tool call. - message.input_image.image_url: Include image urls from the input message. - message.output_text.logprobs: Include logprobs with assistant messages. - reasoning.encrypted_content: Includes an encrypted version of reasoning tokens in reasoning item outputs. This enables reasoning items to be used in multi-turn conversations when using the Responses API statelessly (like when the store parameter is set to false, or when an organization is enrolled in the zero data retention program).	No
input	string or array		No
instructions	string	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
model	string	The model deployment to use for the creation of this response.	Yes
parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
prompt	object	Reference to a prompt template and its variables.	No
└─ id	string	The unique identifier of the prompt template to use.	No
└─ variables	OpenAI.ResponsePromptVariables	Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.	No
└─ version	string	Optional version of the prompt template.	No
reasoning	object	reasoning models only Configuration options for reasoning models.	No
└─ effort	OpenAI.ReasoningEffort	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
└─ generate_summary	enum	Deprecated: use summary instead. A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
└─ summary	enum	A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
store	boolean	Whether to store the generated model response for later retrieval via API.	No	True
stream	boolean	If set to true, the model response data will be streamed to the client as it is generated using server-sent events. See the Streaming section below for more information.	No	False
temperature	number	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. We generally recommend altering this or top_p but not both.	No	1
text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
tool_choice	object	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools.	No
└─ type	OpenAI.ToolChoiceObjectType	Indicates that the model should use a built-in tool to generate a response..	No
tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like file search. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code.	No
top_logprobs	integer	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.	No
top_p	number	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. We generally recommend altering this or temperature but not both.	No	1
truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	AzureResponse
text/event-stream	OpenAI.ResponseStreamEvent

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Examples

Example

Create a model response

POST {endpoint}/openai/v1/responses

Get response

GET {endpoint}/openai/v1/responses/{response_id}

Retrieves a model response with the given ID.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
response_id	path	Yes	string
include_obfuscation	query	No	boolean	When true, stream obfuscation will be enabled. Stream obfuscation adds random characters to an obfuscation field on streaming delta events to normalize payload sizes as a mitigation to certain side-channel attacks. These obfuscation fields are included by default, but add a small amount of overhead to the data stream. You can set include_obfuscation to false to optimize for bandwidth if you trust the network links between your application and the OpenAI API.
include[]	query	No	array

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	AzureResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Delete response

DELETE {endpoint}/openai/v1/responses/{response_id}

Deletes a response by ID.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
response_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	object

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List input items

GET {endpoint}/openai/v1/responses/{response_id}/input_items

Returns a list of input items for a given response.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
response_id	path	Yes	string
limit	query	No	integer	A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
order	query	No	string Possible values: asc, desc	Sort order by the created_at timestamp of the objects. asc for ascending order anddesc for descending order.
after	query	No	string	A cursor for use in pagination. after is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include after=obj_foo in order to fetch the next page of the list.
before	query	No	string	A cursor for use in pagination. before is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include before=obj_foo in order to fetch the previous page of the list.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ResponseItemList

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List vector stores

GET {endpoint}/openai/v1/vector_stores

Returns a list of vector stores.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
limit	query	No	integer	A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
order	query	No	string Possible values: asc, desc	Sort order by the created_at timestamp of the objects. asc for ascending order anddesc for descending order.
after	query	No	string	A cursor for use in pagination. after is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include after=obj_foo in order to fetch the next page of the list.
before	query	No	string	A cursor for use in pagination. before is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include before=obj_foo in order to fetch the previous page of the list.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListVectorStoresResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Createvectorstore

POST {endpoint}/openai/v1/vector_stores

Creates a vector store.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
chunking_strategy	object	The default strategy. This strategy currently uses a max_chunk_size_tokens of 800 and chunk_overlap_tokens of 400.	No
└─ static	OpenAI.StaticChunkingStrategy		No
└─ type	enum	Always static. Possible values: static	No
expires_after	OpenAI.VectorStoreExpirationAfter	The expiration policy for a vector store.	No
file_ids	array	A list of File IDs that the vector store should use. Useful for tools like file_search that can access files.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the vector store.	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Examples

Example file not found: ./examples/vector_stores.json

Get vector store

GET {endpoint}/openai/v1/vector_stores/{vector_store_id}

Retrieves a vector store.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store to retrieve.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Modify vector store

POST {endpoint}/openai/v1/vector_stores/{vector_store_id}

Modifies a vector store.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store to modify.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
expires_after	object	The expiration policy for a vector store.	No
└─ anchor	enum	Anchor timestamp after which the expiration policy applies. Supported anchors: last_active_at. Possible values: last_active_at	No
└─ days	integer	The number of days after the anchor time that the vector store will expire.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the vector store.	No

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Delete vector store

DELETE {endpoint}/openai/v1/vector_stores/{vector_store_id}

Delete a vector store.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store to delete.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.DeleteVectorStoreResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create vector store file batch

POST {endpoint}/openai/v1/vector_stores/{vector_store_id}/file_batches

Create a vector store file batch.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store for which to create a file batch.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	No
chunking_strategy	OpenAI.ChunkingStrategyRequestParam	The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.	No
file_ids	array	A list of File IDs that the vector store should use. Useful for tools like file_search that can access files.	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreFileBatchObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get vector store file batch

GET {endpoint}/openai/v1/vector_stores/{vector_store_id}/file_batches/{batch_id}

Retrieves a vector store file batch.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store that the file batch belongs to.
batch_id	path	Yes	string	The ID of the file batch being retrieved.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreFileBatchObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Cancel vector store file batch

POST {endpoint}/openai/v1/vector_stores/{vector_store_id}/file_batches/{batch_id}/cancel

Cancel a vector store file batch. This attempts to cancel the processing of files in this batch as soon as possible.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store that the file batch belongs to.
batch_id	path	Yes	string	The ID of the file batch to cancel.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreFileBatchObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List files in vector store batch

GET {endpoint}/openai/v1/vector_stores/{vector_store_id}/file_batches/{batch_id}/files

Returns a list of vector store files in a batch.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store that the file batch belongs to.
batch_id	path	Yes	string	The ID of the file batch that the files belong to.
limit	query	No	integer	A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
order	query	No	string Possible values: asc, desc	Sort order by the created_at timestamp of the objects. asc for ascending order anddesc for descending order.
after	query	No	string	A cursor for use in pagination. after is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include after=obj_foo in order to fetch the next page of the list.
before	query	No	string	A cursor for use in pagination. before is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include before=obj_foo in order to fetch the previous page of the list.
filter	query	No		Filter by file status. One of in_progress, completed, failed, cancelled.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListVectorStoreFilesResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

List vector store files

GET {endpoint}/openai/v1/vector_stores/{vector_store_id}/files

Returns a list of vector store files.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store that the files belong to.
limit	query	No	integer	A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
order	query	No	string Possible values: asc, desc	Sort order by the created_at timestamp of the objects. asc for ascending order anddesc for descending order.
after	query	No	string	A cursor for use in pagination. after is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include after=obj_foo in order to fetch the next page of the list.
before	query	No	string	A cursor for use in pagination. before is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include before=obj_foo in order to fetch the previous page of the list.
filter	query	No		Filter by file status. One of in_progress, completed, failed, cancelled.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.ListVectorStoreFilesResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Create vector store file

POST {endpoint}/openai/v1/vector_stores/{vector_store_id}/files

Create a vector store file by attaching a File to a vector store.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store for which to create a File.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	No
chunking_strategy	OpenAI.ChunkingStrategyRequestParam	The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.	No
file_id	string	A File ID that the vector store should use. Useful for tools like file_search that can access files.	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreFileObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Get vector store file

GET {endpoint}/openai/v1/vector_stores/{vector_store_id}/files/{file_id}

Retrieves a vector store file.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store that the file belongs to.
file_id	path	Yes	string	The ID of the file being retrieved.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreFileObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Update vector store file attributes

POST {endpoint}/openai/v1/vector_stores/{vector_store_id}/files/{file_id}

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string
file_id	path	Yes	string

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Request Body

Content-Type: application/json

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	Yes

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.VectorStoreFileObject

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Delete vector store file

DELETE {endpoint}/openai/v1/vector_stores/{vector_store_id}/files/{file_id}

Delete a vector store file. This will remove the file from the vector store but the file itself will not be deleted. To delete the file, use the delete file endpoint.

Parameters

Name	In	Required	Type	Description
endpoint	path	Yes	string url	Supported Azure OpenAI endpoints (protocol and hostname, for example: https://aoairesource.openai.azure.com. Replace “aoairesource” with your Azure OpenAI resource name). https://{your-resource-name}.openai.azure.com
api-version	query	No		The explicit Foundry Models API version to use for this request. v1 if not otherwise specified.
vector_store_id	path	Yes	string	The ID of the vector store that the file belongs to.
file_id	path	Yes	string	The ID of the file to delete.

Request Header

Use either token based authentication or API key. Authenticating with token based authentication is recommended and more secure.

Name	Required	Type	Description
Authorization	True	string	Example: Authorization: Bearer {Azure_OpenAI_Auth_Token} To generate an auth token using Azure CLI: az account get-access-token --resource https://cognitiveservices.azure.com Type: oauth2 Authorization Url: https://login.microsoftonline.com/common/oauth2/v2.0/authorize scope: https://cognitiveservices.azure.com/.default
api-key	True	string	Provide Azure OpenAI API key here

Responses

Status Code: 200

Description: The request has succeeded.

Content-Type	Type	Description
application/json	OpenAI.DeleteVectorStoreFileResponse

Status Code: default

Description: An unexpected error response.

Content-Type	Type	Description
application/json	AzureErrorResponse

Components

AzureAIFoundryModelsApiVersion

Property	Value
Type	string
Values	v1 preview

AzureChatCompletionResponseMessage

The extended response model component for chat completion response messages on the Azure OpenAI service. This model adds support for chat message context, used by the On Your Data feature for intent, citations, and other information related to retrieval-augmented generation performed.

Name	Type	Description	Required	Default
annotations	array	Annotations for the message, when applicable, as when using theweb search tool.	No
audio	object	If the audio output modality is requested, this object contains data about the audio response from the model. .	No
└─ data	string	Base64 encoded audio bytes generated by the model, in the format specified in the request.	No
└─ expires_at	integer	The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.	No
└─ id	string	Unique identifier for this audio response.	No
└─ transcript	string	Transcript of the audio generated by the model.	No
content	string	The contents of the message.	Yes
context	object	An additional property, added to chat completion response messages, produced by the Azure OpenAI service when using extension behavior. This includes intent and citation information from the On Your Data feature.	No
└─ all_retrieved_documents	object	Summary information about documents retrieved by the data retrieval operation.	No
└─ chunk_id	string	The chunk ID for the citation.	No
└─ content	string	The content of the citation.	No
└─ data_source_index	integer	The index of the data source used for retrieval.	No
└─ filepath	string	The file path for the citation.	No
└─ filter_reason	enum	If applicable, an indication of why the document was filtered. Possible values: score, rerank	No
└─ original_search_score	number	The original search score for the retrieval.	No
└─ rerank_score	number	The rerank score for the retrieval.	No
└─ search_queries	array	The search queries executed to retrieve documents.	No
└─ title	string	The title for the citation.	No
└─ url	string	The URL of the citation.	No
└─ citations	array	The citations produced by the data retrieval.	No
└─ intent	string	The detected intent from the chat history, which is used to carry conversation context between interactions	No
function_call	object	Deprecated and replaced by tool_calls. The name and arguments of a function that should be called, as generated by the model.	No
└─ arguments	string		No
└─ name	string		No
reasoning_content	string	An Azure-specific extension property containing generated reasoning content from supported models.	No
refusal	string	The refusal message generated by the model.	Yes
role	enum	The role of the author of this message. Possible values: assistant	Yes
tool_calls	ChatCompletionMessageToolCallsItem	The tool calls generated by the model, such as function calls.	No

AzureChatCompletionStreamResponseDelta

The extended response model for a streaming chat response message on the Azure OpenAI service. This model adds support for chat message context, used by the On Your Data feature for intent, citations, and other information related to retrieval-augmented generation performed.

Name	Type	Description	Required	Default
audio	object		No
└─ data	string		No
└─ expires_at	integer		No
└─ id	string		No
└─ transcript	string		No
content	string	The contents of the chunk message.	No
context	object	An additional property, added to chat completion response messages, produced by the Azure OpenAI service when using extension behavior. This includes intent and citation information from the On Your Data feature.	No
└─ all_retrieved_documents	object	Summary information about documents retrieved by the data retrieval operation.	No
└─ chunk_id	string	The chunk ID for the citation.	No
└─ content	string	The content of the citation.	No
└─ data_source_index	integer	The index of the data source used for retrieval.	No
└─ filepath	string	The file path for the citation.	No
└─ filter_reason	enum	If applicable, an indication of why the document was filtered. Possible values: score, rerank	No
└─ original_search_score	number	The original search score for the retrieval.	No
└─ rerank_score	number	The rerank score for the retrieval.	No
└─ search_queries	array	The search queries executed to retrieve documents.	No
└─ title	string	The title for the citation.	No
└─ url	string	The URL of the citation.	No
└─ citations	array	The citations produced by the data retrieval.	No
└─ intent	string	The detected intent from the chat history, which is used to carry conversation context between interactions	No
function_call	object	Deprecated and replaced by tool_calls. The name and arguments of a function that should be called, as generated by the model.	No
└─ arguments	string		No
└─ name	string		No
reasoning_content	string	An Azure-specific extension property containing generated reasoning content from supported models.	No
refusal	string	The refusal message generated by the model.	No
role	object	The role of the author of a message	No
tool_calls	array		No

AzureChatDataSource

A representation of configuration data for a single Azure OpenAI chat data source. This will be used by a chat completions request that should use Azure OpenAI chat extensions to augment the response behavior. The use of this configuration is compatible only with Azure OpenAI.

Discriminator for AzureChatDataSource

This component uses the property type to discriminate between different types:

Type Value	Schema
azure_search	AzureSearchChatDataSource
azure_cosmos_db	AzureCosmosDBChatDataSource
elasticsearch	ElasticsearchChatDataSource
pinecone	PineconeChatDataSource
mongo_db	MongoDBChatDataSource

Name	Type	Description	Required	Default
type	object		Yes

AzureChatDataSourceAccessTokenAuthenticationOptions

Name	Type	Description	Required	Default
access_token	string		Yes
type	enum	Possible values: access_token	Yes

AzureChatDataSourceApiKeyAuthenticationOptions

Name	Type	Description	Required	Default
key	string		Yes
type	enum	Possible values: api_key	Yes

AzureChatDataSourceAuthenticationOptions

Discriminator for AzureChatDataSourceAuthenticationOptions

This component uses the property type to discriminate between different types:

Type Value	Schema
system_assigned_managed_identity	AzureChatDataSourceSystemAssignedManagedIdentityAuthenticationOptions
user_assigned_managed_identity	AzureChatDataSourceUserAssignedManagedIdentityAuthenticationOptions
access_token	AzureChatDataSourceAccessTokenAuthenticationOptions
connection_string	AzureChatDataSourceConnectionStringAuthenticationOptions
key_and_key_id	AzureChatDataSourceKeyAndKeyIdAuthenticationOptions
encoded_api_key	AzureChatDataSourceEncodedApiKeyAuthenticationOptions
username_and_password	AzureChatDataSourceUsernameAndPasswordAuthenticationOptions

Name	Type	Description	Required	Default
type	AzureChatDataSourceAuthenticationOptionsType		Yes

AzureChatDataSourceAuthenticationOptionsType

Property	Value
Type	string
Values	api_key username_and_password connection_string key_and_key_id encoded_api_key access_token system_assigned_managed_identity user_assigned_managed_identity

AzureChatDataSourceConnectionStringAuthenticationOptions

Name	Type	Description	Required	Default
connection_string	string		Yes
type	enum	Possible values: connection_string	Yes

AzureChatDataSourceDeploymentNameVectorizationSource

Represents a vectorization source that makes internal service calls against an Azure OpenAI embedding model deployment. In contrast with the endpoint-based vectorization source, a deployment-name-based vectorization source must be part of the same Azure OpenAI resource but can be used even in private networks.

Name	Type	Description	Required	Default
deployment_name	string	The embedding model deployment to use for vectorization. This deployment must exist within the same Azure OpenAI resource as the model deployment being used for chat completions.	Yes
dimensions	integer	The number of dimensions to request on embeddings. Only supported in ‘text-embedding-3’ and later models.	No
type	enum	The type identifier, always ‘deployment_name’ for this vectorization source type. Possible values: deployment_name	Yes

AzureChatDataSourceEncodedApiKeyAuthenticationOptions

Name	Type	Description	Required	Default
encoded_api_key	string		Yes
type	enum	Possible values: encoded_api_key	Yes

AzureChatDataSourceEndpointVectorizationSource

Represents a vectorization source that makes public service calls against an Azure OpenAI embedding model deployment.

Name	Type	Description	Required	Default
authentication	object		Yes
└─ access_token	string		No
└─ key	string		No
└─ type	enum	Possible values: access_token	No
dimensions	integer	The number of dimensions to request on embeddings. Only supported in ‘text-embedding-3’ and later models.	No
endpoint	string	Specifies the resource endpoint URL from which embeddings should be retrieved. It should be in the format of: https://YOUR_RESOURCE_NAME.openai.azure.com/openai/deployments/YOUR_DEPLOYMENT_NAME/embeddings. The api-version query parameter is not allowed.	Yes
type	enum	The type identifier, always ‘endpoint’ for this vectorization source type. Possible values: endpoint	Yes

AzureChatDataSourceIntegratedVectorizationSource

Represents an integrated vectorization source as defined within the supporting search resource.

Name	Type	Description	Required	Default
type	enum	The type identifier, always ‘integrated’ for this vectorization source type. Possible values: integrated	Yes

AzureChatDataSourceKeyAndKeyIdAuthenticationOptions

Name	Type	Description	Required	Default
key	string		Yes
key_id	string		Yes
type	enum	Possible values: key_and_key_id	Yes

AzureChatDataSourceModelIdVectorizationSource

Represents a vectorization source that makes service calls based on a search service model ID. This source type is currently only supported by Elasticsearch.

Name	Type	Description	Required	Default
model_id	string	The embedding model build ID to use for vectorization.	Yes
type	enum	The type identifier, always ‘model_id’ for this vectorization source type. Possible values: model_id	Yes

AzureChatDataSourceSystemAssignedManagedIdentityAuthenticationOptions

Name	Type	Description	Required	Default
type	enum	Possible values: system_assigned_managed_identity	Yes

AzureChatDataSourceType

Property	Value
Type	string
Values	azure_search azure_cosmos_db elasticsearch pinecone mongo_db

AzureChatDataSourceUserAssignedManagedIdentityAuthenticationOptions

Name	Type	Description	Required	Default
managed_identity_resource_id	string		Yes
type	enum	Possible values: user_assigned_managed_identity	Yes

AzureChatDataSourceUsernameAndPasswordAuthenticationOptions

Name	Type	Description	Required	Default
password	string		Yes
type	enum	Possible values: username_and_password	Yes
username	string		Yes

AzureChatDataSourceVectorizationSource

A representation of a data vectorization source usable as an embedding resource with a data source.

Discriminator for AzureChatDataSourceVectorizationSource

This component uses the property type to discriminate between different types:

Type Value	Schema
deployment_name	AzureChatDataSourceDeploymentNameVectorizationSource
integrated	AzureChatDataSourceIntegratedVectorizationSource
model_id	AzureChatDataSourceModelIdVectorizationSource

Name	Type	Description	Required	Default
type	object		Yes

AzureChatDataSourceVectorizationSourceType

Property	Value
Type	string
Values	endpoint deployment_name model_id integrated

AzureChatMessageContext

An additional property, added to chat completion response messages, produced by the Azure OpenAI service when using extension behavior. This includes intent and citation information from the On Your Data feature.

Name	Type	Description	Required	Default
all_retrieved_documents	object	Summary information about documents retrieved by the data retrieval operation.	No
└─ chunk_id	string	The chunk ID for the citation.	No
└─ content	string	The content of the citation.	No
└─ data_source_index	integer	The index of the data source used for retrieval.	No
└─ filepath	string	The file path for the citation.	No
└─ filter_reason	enum	If applicable, an indication of why the document was filtered. Possible values: score, rerank	No
└─ original_search_score	number	The original search score for the retrieval.	No
└─ rerank_score	number	The rerank score for the retrieval.	No
└─ search_queries	array	The search queries executed to retrieve documents.	No
└─ title	string	The title for the citation.	No
└─ url	string	The URL of the citation.	No
citations	array	The citations produced by the data retrieval.	No
intent	string	The detected intent from the chat history, which is used to carry conversation context between interactions	No

AzureContentFilterBlocklistResult

A collection of true/false filtering results for configured custom blocklists.

Name	Type	Description	Required	Default
details	array	The pairs of individual blocklist IDs and whether they resulted in a filtering action.	No
filtered	boolean	A value indicating whether any of the detailed blocklists resulted in a filtering action.	Yes

AzureContentFilterCompletionTextSpan

A representation of a span of completion text as used by Azure OpenAI content filter results.

Name	Type	Description	Required	Default
completion_end_offset	integer	Offset of the first UTF32 code point which is excluded from the span. This field is always equal to completion_start_offset for empty spans. This field is always larger than completion_start_offset for non-empty spans.	Yes
completion_start_offset	integer	Offset of the UTF32 code point which begins the span.	Yes

AzureContentFilterCompletionTextSpanDetectionResult

Name	Type	Description	Required	Default
details	array	Detailed information about the detected completion text spans.	Yes
detected	boolean	Whether the labeled content category was detected in the content.	Yes
filtered	boolean	Whether the content detection resulted in a content filtering action.	Yes

AzureContentFilterCustomTopicResult

A collection of true/false filtering results for configured custom topics.

Name	Type	Description	Required	Default
details	array	The pairs of individual topic IDs and whether they are detected.	No
filtered	boolean	A value indicating whether any of the detailed topics resulted in a filtering action.	Yes

AzureContentFilterDetectionResult

A labeled content filter result item that indicates whether the content was detected and whether the content was filtered.

Name	Type	Description	Required	Default
detected	boolean	Whether the labeled content category was detected in the content.	Yes
filtered	boolean	Whether the content detection resulted in a content filtering action.	Yes

AzureContentFilterPersonallyIdentifiableInformationResult

A content filter detection result for Personally Identifiable Information that includes harm extensions.

Name	Type	Description	Required	Default
redacted_text	string	The redacted text with PII information removed or masked.	No
sub_categories	array	Detailed results for individual PIIHarmSubCategory(s).	No

AzureContentFilterResultForChoice

A content filter result for a single response item produced by a generative AI system.

Name	Type	Description	Required	Default
custom_blocklists	object	A collection of true/false filtering results for configured custom blocklists.	No
└─ details	array	The pairs of individual blocklist IDs and whether they resulted in a filtering action.	No
└─ filtered	boolean	A value indicating whether any of the detailed blocklists resulted in a filtering action.	No
custom_topics	object	A collection of true/false filtering results for configured custom topics.	No
└─ details	array	The pairs of individual topic IDs and whether they are detected.	No
└─ filtered	boolean	A value indicating whether any of the detailed topics resulted in a filtering action.	No
error	object	If present, details about an error that prevented content filtering from completing its evaluation.	No
└─ code	integer	A distinct, machine-readable code associated with the error.	No
└─ message	string	A human-readable message associated with the error.	No
hate	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
personally_identifiable_information	object	A content filter detection result for Personally Identifiable Information that includes harm extensions.	No
└─ redacted_text	string	The redacted text with PII information removed or masked.	No
└─ sub_categories	array	Detailed results for individual PIIHarmSubCategory(s).	No
profanity	object	A labeled content filter result item that indicates whether the content was detected and whether the content was filtered.	No
└─ detected	boolean	Whether the labeled content category was detected in the content.	No
└─ filtered	boolean	Whether the content detection resulted in a content filtering action.	No
protected_material_code	object	A detection result that describes a match against licensed code or other protected source material.	No
└─ citation	object	If available, the citation details describing the associated license and its location.	No
└─ URL	string	The URL associated with the license.	No
└─ license	string	The name or identifier of the license associated with the detection.	No
└─ detected	boolean	Whether the labeled content category was detected in the content.	No
└─ filtered	boolean	Whether the content detection resulted in a content filtering action.	No
protected_material_text	object	A labeled content filter result item that indicates whether the content was detected and whether the content was filtered.	No
└─ detected	boolean	Whether the labeled content category was detected in the content.	No
└─ filtered	boolean	Whether the content detection resulted in a content filtering action.	No
self_harm	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
sexual	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
ungrounded_material	AzureContentFilterCompletionTextSpanDetectionResult		No
violence	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No

AzureContentFilterResultForPrompt

A content filter result associated with a single input prompt item into a generative AI system.

Name	Type	Description	Required	Default
content_filter_results	object	The content filter category details for the result.	No
└─ custom_blocklists	object	A collection of true/false filtering results for configured custom blocklists.	No
└─ details	array	The pairs of individual blocklist IDs and whether they resulted in a filtering action.	No
└─ filtered	boolean	A value indicating whether any of the detailed blocklists resulted in a filtering action.	No
└─ custom_topics	object	A collection of true/false filtering results for configured custom topics.	No
└─ details	array	The pairs of individual topic IDs and whether they are detected.	No
└─ filtered	boolean	A value indicating whether any of the detailed topics resulted in a filtering action.	No
└─ error	object	If present, details about an error that prevented content filtering from completing its evaluation.	No
└─ code	integer	A distinct, machine-readable code associated with the error.	No
└─ message	string	A human-readable message associated with the error.	No
└─ hate	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
└─ indirect_attack	object	A labeled content filter result item that indicates whether the content was detected and whether the content was filtered.	No
└─ detected	boolean	Whether the labeled content category was detected in the content.	No
└─ filtered	boolean	Whether the content detection resulted in a content filtering action.	No
└─ jailbreak	object	A labeled content filter result item that indicates whether the content was detected and whether the content was filtered.	No
└─ detected	boolean	Whether the labeled content category was detected in the content.	No
└─ filtered	boolean	Whether the content detection resulted in a content filtering action.	No
└─ profanity	object	A labeled content filter result item that indicates whether the content was detected and whether the content was filtered.	No
└─ detected	boolean	Whether the labeled content category was detected in the content.	No
└─ filtered	boolean	Whether the content detection resulted in a content filtering action.	No
└─ self_harm	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
└─ sexual	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
└─ violence	object	A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.	No
└─ filtered	boolean	Whether the content severity resulted in a content filtering action.	No
└─ severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	No
prompt_index	integer	The index of the input prompt associated with the accompanying content filter result categories.	No

AzureContentFilterSeverityResult

A labeled content filter result item that indicates whether the content was filtered and what the qualitative severity level of the content was, as evaluated against content filter configuration for the category.

Name	Type	Description	Required	Default
filtered	boolean	Whether the content severity resulted in a content filtering action.	Yes
severity	enum	The labeled severity of the content. Possible values: safe, low, medium, high	Yes

AzureCosmosDBChatDataSource

Represents a data source configuration that will use an Azure CosmosDB resource.

Name	Type	Description	Required	Default
parameters	object	The parameter information to control the use of the Azure CosmosDB data source.	Yes
└─ allow_partial_result	boolean	If set to true, the system will allow partial search results to be used and the request will fail if all partial queries fail. If not specified or specified as false, the request will fail if any search query fails.	No	False
└─ authentication	AzureChatDataSourceConnectionStringAuthenticationOptions		No
└─ container_name	string		No
└─ database_name	string		No
└─ embedding_dependency	AzureChatDataSourceVectorizationSource	A representation of a data vectorization source usable as an embedding resource with a data source.	No
└─ fields_mapping	object		No
└─ content_fields	array		No
└─ content_fields_separator	string		No
└─ filepath_field	string		No
└─ title_field	string		No
└─ url_field	string		No
└─ vector_fields	array		No
└─ in_scope	boolean	Whether queries should be restricted to use of the indexed data.	No
└─ include_contexts	array	The output context properties to include on the response. By default, citations and intent will be requested.	No	[‘citations’, ‘intent’]
└─ index_name	string		No
└─ max_search_queries	integer	The maximum number of rewritten queries that should be sent to the search provider for a single user message. By default, the system will make an automatic determination.	No
└─ strictness	integer	The configured strictness of the search relevance filtering. Higher strictness will increase precision but lower recall of the answer.	No
└─ top_n_documents	integer	The configured number of documents to feature in the query.	No
type	enum	The discriminated type identifier, which is always ‘azure_cosmos_db’. Possible values: azure_cosmos_db	Yes

AzureCreateChatCompletionRequest

The extended request model for chat completions against the Azure OpenAI service. This adds the ability to provide data sources for the On Your Data feature.

Name	Type	Description	Required	Default
audio	object	Parameters for audio output. Required when audio output is requested with modalities: ["audio"].	No
└─ format	enum	Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16. Possible values: wav, aac, mp3, flac, opus, pcm16	No
└─ voice	object		No
data_sources	array	The data sources to use for the On Your Data feature, exclusive to Azure OpenAI.	No
frequency_penalty	number	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.	No	0
function_call	enum	Specifying a particular function via {"name": "my_function"} forces the model to call that function. Possible values: none, auto	No
functions	array	Deprecated in favor of tools. A list of functions the model may generate JSON inputs for.	No
logit_bias	object	Modify the likelihood of specified tokens appearing in the completion. Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.	No	None
logprobs	boolean	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.	No	False
max_completion_tokens	integer	An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.	No
max_tokens	integer	The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API. This value is now deprecated in favor of max_completion_tokens, and is not compatible with o1 series models.	No
messages	array	A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text, images, and audio.	Yes
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
modalities	object	Output types that you would like the model to generate. Most models are capable of generating text, which is the default: ["text"] The gpt-4o-audio-preview model can also be used to generate audio. To request that this model generate both text and audio responses, you can use: ["text", "audio"]	No
model	string	The model deployment identifier to use for the chat completion request.	Yes
n	integer	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.	No	1
parallel_tool_calls	object	Whether to enable parallel function calling during tool use.	No
prediction	object	Base representation of predicted output from a model.	No
└─ type	OpenAI.ChatOutputPredictionType		No
presence_penalty	number	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.	No	0
reasoning_effort	object	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
response_format	object		No
└─ type	enum	Possible values: text, json_object, json_schema	No
seed	integer	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.	No
stop	object	Not supported with latest reasoning models o3 and o4-mini. Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.	No
store	boolean	Whether or not to store the output of this chat completion request for use in model distillation or evals products.	No	False
stream	boolean	If set to true, the model response data will be streamed to the client as it is generated using server-sent events.	No	False
stream_options	object	Options for streaming response. Only set this when you set stream: true.	No
└─ include_usage	boolean	If set, an additional chunk will be streamed before the data: [DONE] message. The usage field on this chunk shows the token usage statistics for the entire request, and the choices field will always be an empty array. All other chunks will also include a usage field, but with a null value. NOTE: If the stream is interrupted, you may not receive the final usage chunk which contains the total token usage for the request.	No
temperature	number	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. We generally recommend altering this or top_p but not both.	No	1
tool_choice	OpenAI.ChatCompletionToolChoiceOption	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool. none is the default when no tools are present. auto is the default if tools are present.	No
tools	array	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.	No
top_logprobs	integer	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.	No
top_p	number	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. We generally recommend altering this or temperature but not both.	No	1
user	string	A unique identifier representing your end-user, which can help to monitor and detect abuse.	No
user_security_context	AzureUserSecurityContext	User security context contains several parameters that describe the application itself, and the end user that interacts with the application. These fields assist your security operations teams to investigate and mitigate security incidents by providing a comprehensive approach to protecting your AI applications. Learn more about protecting AI applications using Microsoft Defender for Cloud.	No

AzureCreateChatCompletionResponse

The extended top-level chat completion response model for the Azure OpenAI service. This model adds Responsible AI content filter annotations for prompt input.

Name	Type	Description	Required	Default
choices	array		Yes
created	integer	The Unix timestamp (in seconds) of when the chat completion was created.	Yes
id	string	A unique identifier for the chat completion.	Yes
model	string	The model used for the chat completion.	Yes
object	enum	The object type, which is always chat.completion. Possible values: chat.completion	Yes
prompt_filter_results	array	The Responsible AI content filter annotations associated with prompt inputs into chat completions.	No
system_fingerprint	string	This fingerprint represents the backend configuration that the model runs with. Can be used in conjunction with the seed request parameter to understand when backend changes have been made that might impact determinism.	No
usage	OpenAI.CompletionUsage	Usage statistics for the completion request.	No

AzureCreateChatCompletionStreamResponse

Name	Type	Description	Required	Default
choices	array	A list of chat completion choices. Can contain more than one elements if n is greater than 1. Can also be empty for the last chunk if you set stream_options: {"include_usage": true}.	Yes
content_filter_results	AzureContentFilterResultForChoice	A content filter result for a single response item produced by a generative AI system.	No
created	integer	The Unix timestamp (in seconds) of when the chat completion was created. Each chunk has the same timestamp.	Yes
delta	AzureChatCompletionStreamResponseDelta	The extended response model for a streaming chat response message on the Azure OpenAI service. This model adds support for chat message context, used by the On Your Data feature for intent, citations, and other information related to retrieval-augmented generation performed.	No
id	string	A unique identifier for the chat completion. Each chunk has the same ID.	Yes
model	string	The model to generate the completion.	Yes
object	enum	The object type, which is always chat.completion.chunk. Possible values: chat.completion.chunk	Yes
system_fingerprint	string	This fingerprint represents the backend configuration that the model runs with. Can be used in conjunction with the seed request parameter to understand when backend changes have been made that might impact determinism.	No
usage	object	Usage statistics for the completion request.	No
└─ completion_tokens	integer	Number of tokens in the generated completion.	No	0
└─ completion_tokens_details	object	Breakdown of tokens used in a completion.	No
└─ accepted_prediction_tokens	integer	When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.	No	0
└─ audio_tokens	integer	Audio input tokens generated by the model.	No	0
└─ reasoning_tokens	integer	Tokens generated by the model for reasoning.	No	0
└─ rejected_prediction_tokens	integer	When using Predicted Outputs, the number of tokens in the prediction that did not appear in the completion. However, like reasoning tokens, these tokens are still counted in the total completion tokens for purposes of billing, output, and context window limits.	No	0
└─ prompt_tokens	integer	Number of tokens in the prompt.	No	0
└─ prompt_tokens_details	object	Breakdown of tokens used in the prompt.	No
└─ audio_tokens	integer	Audio input tokens present in the prompt.	No	0
└─ cached_tokens	integer	Cached tokens present in the prompt.	No	0
└─ total_tokens	integer	Total number of tokens used in the request (prompt + completion).	No	0

AzureCreateEmbeddingRequest

Name	Type	Description	Required	Default
dimensions	integer	The number of dimensions the resulting output embeddings should have. Only supported in text-embedding-3 and later models.	No
encoding_format	enum	The format to return the embeddings in. Can be either float or base64. Possible values: float, base64	No
input	string or array		Yes
model	string	The model to use for the embedding request.	Yes
user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No

AzureCreateFileRequestMultiPart

Name	Type	Description	Required	Default
expires_after	object		Yes
└─ anchor	AzureFileExpiryAnchor		No
└─ seconds	integer		No
file	string		Yes
purpose	enum	The intended purpose of the uploaded file. One of: - assistants: Used in the Assistants API - batch: Used in the Batch API - fine-tune: Used for fine-tuning - evals: Used for eval data sets Possible values: assistants, batch, fine-tune, evals	Yes

AzureCreateResponse

Name	Type	Description	Required	Default
background	boolean	Whether to run the model response in the background. Learn more.	No	False
include	array	Specify additional output data to include in the model response. Currently supported values are: - code_interpreter_call.outputs: Includes the outputs of python code execution in code interpreter tool call items. - computer_call_output.output.image_url: Include image urls from the computer call output. - file_search_call.results: Include the search results of the file search tool call. - message.input_image.image_url: Include image urls from the input message. - message.output_text.logprobs: Include logprobs with assistant messages. - reasoning.encrypted_content: Includes an encrypted version of reasoning tokens in reasoning item outputs. This enables reasoning items to be used in multi-turn conversations when using the Responses API statelessly (like when the store parameter is set to false, or when an organization is enrolled in the zero data retention program).	No
input	string or array		No
instructions	string	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
model	string	The model deployment to use for the creation of this response.	Yes
parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
prompt	object	Reference to a prompt template and its variables.	No
└─ id	string	The unique identifier of the prompt template to use.	No
└─ variables	OpenAI.ResponsePromptVariables	Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.	No
└─ version	string	Optional version of the prompt template.	No
reasoning	object	reasoning models only Configuration options for reasoning models.	No
└─ effort	OpenAI.ReasoningEffort	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
└─ generate_summary	enum	Deprecated: use summary instead. A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
└─ summary	enum	A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
store	boolean	Whether to store the generated model response for later retrieval via API.	No	True
stream	boolean	If set to true, the model response data will be streamed to the client as it is generated using server-sent events. See the Streaming section below for more information.	No	False
temperature	number	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. We generally recommend altering this or top_p but not both.	No	1
text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
tool_choice	object	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools.	No
└─ type	OpenAI.ToolChoiceObjectType	Indicates that the model should use a built-in tool to generate a response..	No
tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like file search. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code.	No
top_logprobs	integer	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.	No
top_p	number	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. We generally recommend altering this or temperature but not both.	No	1
truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No

AzureErrorResponse

Name	Type	Description	Required	Default
error	object	The error details.	No
└─ code	string	The distinct, machine-generated identifier for the error.	No
└─ inner_error			No
└─ message	string	A human-readable message associated with the error.	No
└─ param	string	If applicable, the request input parameter associated with the error	No
└─ type	enum	The object type, always ‘error.‘ Possible values: error	No

AzureEvalAPICompletionsSamplingParams

Name	Type	Description	Required	Default
parallel_tool_calls	boolean		No
response_format	OpenAI.ResponseTextFormatConfiguration		No
tools	array		No

AzureEvalAPIModelSamplingParams

Name	Type	Description	Required	Default
max_tokens	integer	The maximum number of tokens in the generated output.	No
reasoning_effort	enum	Controls the level of reasoning effort applied during generation. Possible values: low, medium, high	No
seed	integer	A seed value to initialize the randomness during sampling.	No
temperature	number	A higher temperature increases randomness in the outputs.	No
top_p	number	An alternative to temperature for nucleus sampling; 1.0 includes all tokens.	No

AzureEvalAPIResponseSamplingParams

Name	Type	Description	Required	Default
parallel_tool_calls	boolean		No
response_format	OpenAI.ResponseTextFormatConfiguration		No
tools	array		No

AzureFileExpiryAnchor

Property	Value
Type	string
Values	created_at

AzureFineTuneReinforcementMethod

Name	Type	Description	Required	Default
grader	object	A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.	Yes
└─ calculate_output	string	A formula to calculate the output based on grader results.	No
└─ evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	No
└─ graders	object		No
└─ input	array	The input text. This may include template strings.	No
└─ model	string	The model to use for the evaluation.	No
└─ name	string	The name of the grader.	No
└─ operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	No
└─ range	array	The range of the score. Defaults to [0, 1].	No
└─ reference	string	The text being graded against.	No
└─ sampling_params		The sampling parameters for the model.	No
└─ type	enum	The object type, which is always multi. Possible values: multi	No
hyperparameters	OpenAI.FineTuneReinforcementHyperparameters	The hyperparameters used for the reinforcement fine-tuning job.	No
response_format	object		No
└─ json_schema	object	JSON Schema for the response format	No
└─ type	enum	Type of response format Possible values: json_schema	No

AzureListFilesResponse

Name	Type	Description	Required	Default
data	array		Yes
first_id	string		Yes
has_more	boolean		Yes
last_id	string		Yes
object	enum	Possible values: list	Yes

AzureOpenAIFile

Name	Type	Description	Required	Default
bytes	integer	The size of the file, in bytes.	Yes
created_at	integer	The Unix timestamp (in seconds) for when the file was created.	Yes
expires_at	integer	The Unix timestamp (in seconds) for when the file will expire.	No
filename	string	The name of the file.	Yes
id	string	The file identifier, which can be referenced in the API endpoints.	Yes
object	enum	The object type, which is always file. Possible values: file	Yes
purpose	enum	The intended purpose of the file. Supported values are assistants, assistants_output, batch, batch_output, fine-tune and fine-tune-results. Possible values: assistants, assistants_output, batch, batch_output, fine-tune, fine-tune-results, evals	Yes
status	enum	Possible values: uploaded, pending, running, processed, error, deleting, deleted	Yes
status_details	string	Deprecated. For details on why a fine-tuning training file failed validation, see the error field on fine_tuning.job.	No

AzurePiiSubCategoryResult

Result details for individual PIIHarmSubCategory(s).

Name	Type	Description	Required	Default
detected	boolean	Whether the labeled content subcategory was detected in the content.	Yes
filtered	boolean	Whether the content detection resulted in a content filtering action for this subcategory.	Yes
redacted	boolean	Whether the content was redacted for this subcategory.	Yes
sub_category	string	The PIIHarmSubCategory that was evaluated.	Yes

AzureResponse

Name	Type	Description	Required	Default
background	boolean	Whether to run the model response in the background. Learn more.	No	False
created_at	integer	Unix timestamp (in seconds) of when this Response was created.	Yes
error	object	An error object returned when the model fails to generate a Response.	Yes
└─ code	OpenAI.ResponseErrorCode	The error code for the response.	No
└─ message	string	A human-readable description of the error.	No
id	string	Unique identifier for this Response.	Yes
incomplete_details	object	Details about why the response is incomplete.	Yes
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
instructions	string or array		Yes
max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
model	string	The model used to generate this response.	Yes
object	enum	The object type of this resource - always set to response. Possible values: response	Yes
output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	Yes
output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	Yes	True
previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
prompt	object	Reference to a prompt template and its variables.	No
└─ id	string	The unique identifier of the prompt template to use.	No
└─ variables	OpenAI.ResponsePromptVariables	Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.	No
└─ version	string	Optional version of the prompt template.	No
reasoning	object	reasoning models only Configuration options for reasoning models.	No
└─ effort	OpenAI.ReasoningEffort	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
└─ generate_summary	enum	Deprecated: use summary instead. A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
└─ summary	enum	A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
temperature	number	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. We generally recommend altering this or top_p but not both.	Yes
text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
tool_choice	object	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools.	No
└─ type	OpenAI.ToolChoiceObjectType	Indicates that the model should use a built-in tool to generate a response..	No
tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
top_logprobs	integer	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.	No
top_p	number	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. We generally recommend altering this or temperature but not both.	Yes
truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	Yes

AzureSearchChatDataSource

Represents a data source configuration that will use an Azure Search resource.

Name	Type	Description	Required	Default
parameters	object	The parameter information to control the use of the Azure Search data source.	Yes
└─ allow_partial_result	boolean	If set to true, the system will allow partial search results to be used and the request will fail if all partial queries fail. If not specified or specified as false, the request will fail if any search query fails.	No	False
└─ authentication	object		No
└─ access_token	string		No
└─ key	string		No
└─ managed_identity_resource_id	string		No
└─ type	enum	Possible values: access_token	No
└─ embedding_dependency	object	Represents a vectorization source that makes public service calls against an Azure OpenAI embedding model deployment.	No
└─ authentication	AzureChatDataSourceApiKeyAuthenticationOptions or AzureChatDataSourceAccessTokenAuthenticationOptions	The authentication mechanism to use with the endpoint-based vectorization source. Endpoint authentication supports API key and access token mechanisms.	No
└─ deployment_name	string	The embedding model deployment to use for vectorization. This deployment must exist within the same Azure OpenAI resource as the model deployment being used for chat completions.	No
└─ dimensions	integer	The number of dimensions to request on embeddings. Only supported in ‘text-embedding-3’ and later models.	No
└─ endpoint	string	Specifies the resource endpoint URL from which embeddings should be retrieved. It should be in the format of: https://YOUR_RESOURCE_NAME.openai.azure.com/openai/deployments/YOUR_DEPLOYMENT_NAME/embeddings. The api-version query parameter is not allowed.	No
└─ type	enum	The type identifier, always ‘integrated’ for this vectorization source type. Possible values: integrated	No
└─ endpoint	string	The absolute endpoint path for the Azure Search resource to use.	No
└─ fields_mapping	object	The field mappings to use with the Azure Search resource.	No
└─ content_fields	array	The names of index fields that should be treated as content.	No
└─ content_fields_separator	string	The separator pattern that content fields should use.	No
└─ filepath_field	string	The name of the index field to use as a filepath.	No
└─ image_vector_fields	array	The names of fields that represent image vector data.	No
└─ title_field	string	The name of the index field to use as a title.	No
└─ url_field	string	The name of the index field to use as a URL.	No
└─ vector_fields	array	The names of fields that represent vector data.	No
└─ filter	string	A filter to apply to the search.	No
└─ in_scope	boolean	Whether queries should be restricted to use of the indexed data.	No
└─ include_contexts	array	The output context properties to include on the response. By default, citations and intent will be requested.	No	[‘citations’, ‘intent’]
└─ index_name	string	The name of the index to use, as specified in the Azure Search resource.	No
└─ max_search_queries	integer	The maximum number of rewritten queries that should be sent to the search provider for a single user message. By default, the system will make an automatic determination.	No
└─ query_type	enum	The query type for the Azure Search resource to use. Possible values: simple, semantic, vector, vector_simple_hybrid, vector_semantic_hybrid	No
└─ semantic_configuration	string	Additional semantic configuration for the query.	No
└─ strictness	integer	The configured strictness of the search relevance filtering. Higher strictness will increase precision but lower recall of the answer.	No
└─ top_n_documents	integer	The configured number of documents to feature in the query.	No
type	enum	The discriminated type identifier, which is always ‘azure_search’. Possible values: azure_search	Yes

AzureUserSecurityContext

User security context contains several parameters that describe the application itself, and the end user that interacts with the application. These fields assist your security operations teams to investigate and mitigate security incidents by providing a comprehensive approach to protecting your AI applications. Learn more about protecting AI applications using Microsoft Defender for Cloud.

Name	Type	Description	Required	Default
application_name	string	The name of the application. Sensitive personal information should not be included in this field.	No
end_user_id	string	This identifier is the Microsoft Entra ID (formerly Azure Active Directory) user object ID used to authenticate end-users within the generative AI application. Sensitive personal information should not be included in this field.	No
end_user_tenant_id	string	The Microsoft 365 tenant ID the end user belongs to. It’s required when the generative AI application is multitenant.	No
source_ip	string	Captures the original client’s IP address.	No

ChatCompletionMessageToolCallsItem

The tool calls generated by the model, such as function calls.

Array of: OpenAI.ChatCompletionMessageToolCall

CopiedAccountDetails

Name	Type	Description	Required	Default
destinationResourceId	string	The ID of the destination resource where the model was copied to.	Yes
region	string	The region where the model was copied to.	Yes
status	enum	The status of the copy operation. Possible values: Completed, Failed, InProgress	Yes

CopyModelRequest

Name	Type	Description	Required	Default
destinationResourceId	string	The ID of the destination Resource to copy.	Yes
region	string	The region to copy the model to.	Yes

CopyModelResponse

Name	Type	Description	Required	Default
checkpointedModelName	string	The ID of the copied model.	Yes
copiedAccountDetails	array	The ID of the destination resource id where it was copied	Yes
fineTuningJobId	string	The ID of the fine-tuning job that the checkpoint was copied from.	Yes

ElasticsearchChatDataSource

Name	Type	Description	Required	Default
parameters	object	The parameter information to control the use of the Elasticsearch data source.	Yes
└─ allow_partial_result	boolean	If set to true, the system will allow partial search results to be used and the request will fail if all partial queries fail. If not specified or specified as false, the request will fail if any search query fails.	No	False
└─ authentication	object		No
└─ encoded_api_key	string		No
└─ key	string		No
└─ key_id	string		No
└─ type	enum	Possible values: encoded_api_key	No
└─ embedding_dependency	AzureChatDataSourceVectorizationSource	A representation of a data vectorization source usable as an embedding resource with a data source.	No
└─ endpoint	string		No
└─ fields_mapping	object		No
└─ content_fields	array		No
└─ content_fields_separator	string		No
└─ filepath_field	string		No
└─ title_field	string		No
└─ url_field	string		No
└─ vector_fields	array		No
└─ in_scope	boolean	Whether queries should be restricted to use of the indexed data.	No
└─ include_contexts	array	The output context properties to include on the response. By default, citations and intent will be requested.	No	[‘citations’, ‘intent’]
└─ index_name	string		No
└─ max_search_queries	integer	The maximum number of rewritten queries that should be sent to the search provider for a single user message. By default, the system will make an automatic determination.	No
└─ query_type	enum	Possible values: simple, vector	No
└─ strictness	integer	The configured strictness of the search relevance filtering. Higher strictness will increase precision but lower recall of the answer.	No
└─ top_n_documents	integer	The configured number of documents to feature in the query.	No
type	enum	The discriminated type identifier, which is always ‘elasticsearch’. Possible values: elasticsearch	Yes

MongoDBChatDataSource

Name	Type	Description	Required	Default
parameters	object	The parameter information to control the use of the MongoDB data source.	Yes
└─ allow_partial_result	boolean	If set to true, the system will allow partial search results to be used and the request will fail if all partial queries fail. If not specified or specified as false, the request will fail if any search query fails.	No	False
└─ app_name	string	The name of the MongoDB application.	No
└─ authentication	object		No
└─ password	string		No
└─ type	enum	Possible values: username_and_password	No
└─ username	string		No
└─ collection_name	string	The name of the MongoDB collection.	No
└─ database_name	string	The name of the MongoDB database.	No
└─ embedding_dependency	object	Represents a vectorization source that makes public service calls against an Azure OpenAI embedding model deployment.	No
└─ authentication	AzureChatDataSourceApiKeyAuthenticationOptions or AzureChatDataSourceAccessTokenAuthenticationOptions	The authentication mechanism to use with the endpoint-based vectorization source. Endpoint authentication supports API key and access token mechanisms.	No
└─ deployment_name	string	The embedding model deployment to use for vectorization. This deployment must exist within the same Azure OpenAI resource as the model deployment being used for chat completions.	No
└─ dimensions	integer	The number of dimensions to request on embeddings. Only supported in ‘text-embedding-3’ and later models.	No
└─ endpoint	string	Specifies the resource endpoint URL from which embeddings should be retrieved. It should be in the format of: https://YOUR_RESOURCE_NAME.openai.azure.com/openai/deployments/YOUR_DEPLOYMENT_NAME/embeddings. The api-version query parameter is not allowed.	No
└─ type	enum	The type identifier, always ‘deployment_name’ for this vectorization source type. Possible values: deployment_name	No
└─ endpoint	string	The name of the MongoDB cluster endpoint.	No
└─ fields_mapping	object	Field mappings to apply to data used by the MongoDB data source. Note that content and vector field mappings are required for MongoDB.	No
└─ content_fields	array		No
└─ content_fields_separator	string		No
└─ filepath_field	string		No
└─ title_field	string		No
└─ url_field	string		No
└─ vector_fields	array		No
└─ in_scope	boolean	Whether queries should be restricted to use of the indexed data.	No
└─ include_contexts	array	The output context properties to include on the response. By default, citations and intent will be requested.	No	[‘citations’, ‘intent’]
└─ index_name	string	The name of the MongoDB index.	No
└─ max_search_queries	integer	The maximum number of rewritten queries that should be sent to the search provider for a single user message. By default, the system will make an automatic determination.	No
└─ strictness	integer	The configured strictness of the search relevance filtering. Higher strictness will increase precision but lower recall of the answer.	No
└─ top_n_documents	integer	The configured number of documents to feature in the query.	No
type	enum	The discriminated type identifier, which is always ‘mongo_db’. Possible values: mongo_db	Yes

OpenAI.Annotation

Discriminator for OpenAI.Annotation

This component uses the property type to discriminate between different types:

Type Value	Schema
file_citation	OpenAI.AnnotationFileCitation
url_citation	OpenAI.AnnotationUrlCitation
file_path	OpenAI.AnnotationFilePath

Name	Type	Description	Required	Default
type	OpenAI.AnnotationType		Yes

OpenAI.AnnotationFileCitation

A citation to a file.

Name	Type	Description	Required	Default
file_id	string	The ID of the file.	Yes
filename	string	The filename of the file cited.	Yes
index	integer	The index of the file in the list of files.	Yes
type	enum	The type of the file citation. Always file_citation. Possible values: file_citation	Yes

OpenAI.AnnotationFilePath

A path to a file.

Name	Type	Description	Required	Default
file_id	string	The ID of the file.	Yes
index	integer	The index of the file in the list of files.	Yes
type	enum	The type of the file path. Always file_path. Possible values: file_path	Yes

OpenAI.AnnotationType

Property	Value
Type	string
Values	file_citation url_citation file_path container_file_citation

OpenAI.AnnotationUrlCitation

A citation for a web resource used to generate a model response.

Name	Type	Description	Required	Default
end_index	integer	The index of the last character of the URL citation in the message.	Yes
start_index	integer	The index of the first character of the URL citation in the message.	Yes
title	string	The title of the web resource.	Yes
type	enum	The type of the URL citation. Always url_citation. Possible values: url_citation	Yes
url	string	The URL of the web resource.	Yes

OpenAI.ApproximateLocation

Name	Type	Description	Required	Default
city	string		No
country	string		No
region	string		No
timezone	string		No
type	enum	Possible values: approximate	Yes

OpenAI.AutoChunkingStrategyRequestParam

The default strategy. This strategy currently uses a max_chunk_size_tokens of 800 and chunk_overlap_tokens of 400.

Name	Type	Description	Required	Default
type	enum	Always auto. Possible values: auto	Yes

OpenAI.ChatCompletionFunctionCallOption

Specifying a particular function via {"name": "my_function"} forces the model to call that function.

Name	Type	Description	Required	Default
name	string	The name of the function to call.	Yes

OpenAI.ChatCompletionFunctions

Name	Type	Description	Required	Default
description	string	A description of what the function does, used by the model to choose when and how to call the function.	No
name	string	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.	Yes
parameters		The parameters the functions accepts, described as a JSON Schema object. See the JSON Schema reference for documentation about the format. Omitting parameters defines a function with an empty parameter list.	No

OpenAI.ChatCompletionMessageAudioChunk

Name	Type	Description	Required	Default
data	string		No
expires_at	integer		No
id	string		No
transcript	string		No

OpenAI.ChatCompletionMessageToolCall

Name	Type	Description	Required	Default
function	object	The function that the model called.	Yes
└─ arguments	string	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.	No
└─ name	string	The name of the function to call.	No
id	string	The ID of the tool call.	Yes
type	enum	The type of the tool. Currently, only function is supported. Possible values: function	Yes

OpenAI.ChatCompletionMessageToolCallChunk

Name	Type	Description	Required	Default
function	object		No
└─ arguments	string	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.	No
└─ name	string	The name of the function to call.	No
id	string	The ID of the tool call.	No
index	integer		Yes
type	enum	The type of the tool. Currently, only function is supported. Possible values: function	No

OpenAI.ChatCompletionNamedToolChoice

Specifies a tool the model should use. Use to force the model to call a specific function.

Name	Type	Description	Required	Default
function	object		Yes
└─ name	string	The name of the function to call.	No
type	enum	The type of the tool. Currently, only function is supported. Possible values: function	Yes

OpenAI.ChatCompletionRequestAssistantMessage

Messages sent by the model in response to user messages.

Name	Type	Description	Required	Default
audio	object	Data about a previous audio response from the model.	No
└─ id	string	Unique identifier for a previous audio response from the model.	No
content	string or array		No
function_call	object	Deprecated and replaced by tool_calls. The name and arguments of a function that should be called, as generated by the model.	No
└─ arguments	string		No
└─ name	string		No
name	string	An optional name for the participant. Provides the model information to differentiate between participants of the same role.	No
refusal	string	The refusal message by the assistant.	No
role	enum	The role of the messages author, in this case assistant. Possible values: assistant	Yes
tool_calls	ChatCompletionMessageToolCallsItem	The tool calls generated by the model, such as function calls.	No

OpenAI.ChatCompletionRequestAssistantMessageContentPart

Name	Type	Description	Required	Default
refusal	string	The refusal message generated by the model.	Yes
text	string	The text content.	Yes
type	enum	The type of the content part. Possible values: refusal	Yes

OpenAI.ChatCompletionRequestDeveloperMessage

Developer-provided instructions that the model should follow, regardless of messages sent by the user. With o1 models and newer, developer messages replace the previous system messages.

Name	Type	Description	Required	Default
content	string or array		Yes
name	string	An optional name for the participant. Provides the model information to differentiate between participants of the same role.	No
role	enum	The role of the messages author, in this case developer. Possible values: developer	Yes

OpenAI.ChatCompletionRequestFunctionMessage

Name	Type	Description	Required	Default
content	string	The contents of the function message.	Yes
name	string	The name of the function to call.	Yes
role	enum	The role of the messages author, in this case function. Possible values: function	Yes

OpenAI.ChatCompletionRequestMessage

Discriminator for OpenAI.ChatCompletionRequestMessage

This component uses the property role to discriminate between different types:

Type Value	Schema
system	OpenAI.ChatCompletionRequestSystemMessage
developer	OpenAI.ChatCompletionRequestDeveloperMessage
user	OpenAI.ChatCompletionRequestUserMessage
assistant	OpenAI.ChatCompletionRequestAssistantMessage
tool	OpenAI.ChatCompletionRequestToolMessage
function	OpenAI.ChatCompletionRequestFunctionMessage

Name	Type	Description	Required	Default
content	string or array		No
role	object	The role of the author of a message	Yes

OpenAI.ChatCompletionRequestMessageContentPart

Discriminator for OpenAI.ChatCompletionRequestMessageContentPart

This component uses the property type to discriminate between different types:

Type Value	Schema
text	OpenAI.ChatCompletionRequestMessageContentPartText
image_url	OpenAI.ChatCompletionRequestMessageContentPartImage
refusal	OpenAI.ChatCompletionRequestMessageContentPartRefusal
file	OpenAI.ChatCompletionRequestMessageContentPartFile
input_audio	OpenAI.ChatCompletionRequestMessageContentPartAudio

Name	Type	Description	Required	Default
type	OpenAI.ChatCompletionRequestMessageContentPartType		Yes

OpenAI.ChatCompletionRequestMessageContentPartAudio

Name	Type	Description	Required	Default
input_audio	object		Yes
└─ data	string	Base64 encoded audio data.	No
└─ format	enum	The format of the encoded audio data. Currently supports “wav” and “mp3”. Possible values: wav, mp3	No
type	enum	The type of the content part. Always input_audio. Possible values: input_audio	Yes

OpenAI.ChatCompletionRequestMessageContentPartFile

Name	Type	Description	Required	Default
file	object		Yes
└─ file_data	string	The base64 encoded file data, used when passing the file to the model as a string.	No
└─ file_id	string	The ID of an uploaded file to use as input.	No
└─ filename	string	The name of the file, used when passing the file to the model as a string.	No
type	enum	The type of the content part. Always file. Possible values: file	Yes

OpenAI.ChatCompletionRequestMessageContentPartImage

Name	Type	Description	Required	Default
image_url	object		Yes
└─ detail	enum	Specifies the detail level of the image. Possible values: auto, low, high	No
└─ url	string	Either a URL of the image or the base64 encoded image data.	No
type	enum	The type of the content part. Possible values: image_url	Yes

OpenAI.ChatCompletionRequestMessageContentPartRefusal

Name	Type	Description	Required	Default
refusal	string	The refusal message generated by the model.	Yes
type	enum	The type of the content part. Possible values: refusal	Yes

OpenAI.ChatCompletionRequestMessageContentPartText

Name	Type	Description	Required	Default
text	string	The text content.	Yes
type	enum	The type of the content part. Possible values: text	Yes

OpenAI.ChatCompletionRequestMessageContentPartType

Property	Value
Type	string
Values	text file input_audio image_url refusal

OpenAI.ChatCompletionRequestSystemMessage

Developer-provided instructions that the model should follow, regardless of messages sent by the user. With o1 models and newer, use developer messages for this purpose instead.

Name	Type	Description	Required	Default
content	string or array		Yes
name	string	An optional name for the participant. Provides the model information to differentiate between participants of the same role.	No
role	enum	The role of the messages author, in this case system. Possible values: system	Yes

OpenAI.ChatCompletionRequestSystemMessageContentPart

References: OpenAI.ChatCompletionRequestMessageContentPartText

OpenAI.ChatCompletionRequestToolMessage

Name	Type	Description	Required	Default
content	string or array		Yes
role	enum	The role of the messages author, in this case tool. Possible values: tool	Yes
tool_call_id	string	Tool call that this message is responding to.	Yes

OpenAI.ChatCompletionRequestToolMessageContentPart

References: OpenAI.ChatCompletionRequestMessageContentPartText

OpenAI.ChatCompletionRequestUserMessage

Messages sent by an end user, containing prompts or additional context information.

Name	Type	Description	Required	Default
content	string or array		Yes
name	string	An optional name for the participant. Provides the model information to differentiate between participants of the same role.	No
role	enum	The role of the messages author, in this case user. Possible values: user	Yes

OpenAI.ChatCompletionRequestUserMessageContentPart

Name	Type	Description	Required	Default
file	object		Yes
└─ file_data	string	The base64 encoded file data, used when passing the file to the model as a string.	No
└─ file_id	string	The ID of an uploaded file to use as input.	No
└─ filename	string	The name of the file, used when passing the file to the model as a string.	No
image_url	object		Yes
└─ detail	enum	Specifies the detail level of the image. Possible values: auto, low, high	No
└─ url	string	Either a URL of the image or the base64 encoded image data.	No
input_audio	object		Yes
└─ data	string	Base64 encoded audio data.	No
└─ format	enum	The format of the encoded audio data. Currently supports “wav” and “mp3”. Possible values: wav, mp3	No
text	string	The text content.	Yes
type	enum	The type of the content part. Always file. Possible values: file	Yes

OpenAI.ChatCompletionRole

The role of the author of a message

Property	Value
Description	The role of the author of a message
Type	string
Values	system developer user assistant tool function

OpenAI.ChatCompletionStreamOptions

Options for streaming response. Only set this when you set stream: true.

Name	Type	Description	Required	Default
include_usage	boolean	If set, an additional chunk will be streamed before the data: [DONE] message. The usage field on this chunk shows the token usage statistics for the entire request, and the choices field will always be an empty array. All other chunks will also include a usage field, but with a null value. NOTE: If the stream is interrupted, you may not receive the final usage chunk which contains the total token usage for the request.	No

OpenAI.ChatCompletionStreamResponseDelta

A chat completion delta generated by streamed model responses.

Name	Type	Description	Required	Default
audio	object		No
└─ data	string		No
└─ expires_at	integer		No
└─ id	string		No
└─ transcript	string		No
content	string	The contents of the chunk message.	No
function_call	object	Deprecated and replaced by tool_calls. The name and arguments of a function that should be called, as generated by the model.	No
└─ arguments	string		No
└─ name	string		No
refusal	string	The refusal message generated by the model.	No
role	object	The role of the author of a message	No
tool_calls	array		No

OpenAI.ChatCompletionTokenLogprob

Name	Type	Description	Required	Default
bytes	array	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.	Yes
logprob	number	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.	Yes
token	string	The token.	Yes
top_logprobs	array	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.	Yes

OpenAI.ChatCompletionTool

Name	Type	Description	Required	Default
function	OpenAI.FunctionObject		Yes
type	enum	The type of the tool. Currently, only function is supported. Possible values: function	Yes

OpenAI.ChatCompletionToolChoiceOption

Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool.

none is the default when no tools are present. auto is the default if tools are present.

Name	Type	Description	Required	Default
function	object		Yes
└─ name	string	The name of the function to call.	No
type	enum	The type of the tool. Currently, only function is supported. Possible values: function	Yes

OpenAI.ChatOutputPrediction

Base representation of predicted output from a model.

Discriminator for OpenAI.ChatOutputPrediction

This component uses the property type to discriminate between different types:

Type Value	Schema
content	OpenAI.ChatOutputPredictionContent

Name	Type	Description	Required	Default
type	OpenAI.ChatOutputPredictionType		Yes

OpenAI.ChatOutputPredictionContent

Static predicted output content, such as the content of a text file that is being regenerated.

Name	Type	Description	Required	Default
content	string or array		Yes
type	enum	The type of the predicted content you want to provide. This type is currently always content. Possible values: content	Yes

OpenAI.ChatOutputPredictionType

Property	Value
Type	string
Values	content

OpenAI.ChunkingStrategyRequestParam

The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.

Discriminator for OpenAI.ChunkingStrategyRequestParam

This component uses the property type to discriminate between different types:

Type Value	Schema
static	OpenAI.StaticChunkingStrategyRequestParam

Name	Type	Description	Required	Default
type	enum	The type of chunking strategy. Possible values: auto, static	Yes

OpenAI.ChunkingStrategyResponseParam

Discriminator for OpenAI.ChunkingStrategyResponseParam

This component uses the property type to discriminate between different types:

Type Value	Schema
other	OpenAI.OtherChunkingStrategyResponseParam
static	OpenAI.StaticChunkingStrategyResponseParam

Name	Type	Description	Required	Default
type	enum	Possible values: static, other	Yes

OpenAI.CodeInterpreterOutput

Discriminator for OpenAI.CodeInterpreterOutput

This component uses the property type to discriminate between different types:

Type Value	Schema
image	OpenAI.CodeInterpreterOutputImage
logs	OpenAI.CodeInterpreterOutputLogs

Name	Type	Description	Required	Default
type	OpenAI.CodeInterpreterOutputType		Yes

OpenAI.CodeInterpreterOutputImage

The image output from the code interpreter.

Name	Type	Description	Required	Default
type	enum	The type of the output. Always ‘image’. Possible values: image	Yes
url	string	The URL of the image output from the code interpreter.	Yes

OpenAI.CodeInterpreterOutputLogs

The logs output from the code interpreter.

Name	Type	Description	Required	Default
logs	string	The logs output from the code interpreter.	Yes
type	enum	The type of the output. Always ‘logs’. Possible values: logs	Yes

OpenAI.CodeInterpreterOutputType

Property	Value
Type	string
Values	logs image

OpenAI.CodeInterpreterTool

A tool that runs Python code to help generate a response to a prompt.

Name	Type	Description	Required	Default
container	object	Configuration for a code interpreter container. Optionally specify the IDs of the files to run the code on.	Yes
└─ file_ids	array	An optional list of uploaded files to make available to your code.	No
└─ type	enum	Always auto. Possible values: auto	No
type	enum	The type of the code interpreter tool. Always code_interpreter. Possible values: code_interpreter	Yes

OpenAI.CodeInterpreterToolAuto

Configuration for a code interpreter container. Optionally specify the IDs of the files to run the code on.

Name	Type	Description	Required	Default
file_ids	array	An optional list of uploaded files to make available to your code.	No
type	enum	Always auto. Possible values: auto	Yes

OpenAI.CodeInterpreterToolCallItemParam

A tool call to run code.

Name	Type	Description	Required	Default
code	string	The code to run, or null if not available.	Yes
container_id	string	The ID of the container used to run the code.	Yes
outputs	array	The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available.	Yes
type	enum	Possible values: code_interpreter_call	Yes

OpenAI.CodeInterpreterToolCallItemResource

A tool call to run code.

Name	Type	Description	Required	Default
code	string	The code to run, or null if not available.	Yes
container_id	string	The ID of the container used to run the code.	Yes
outputs	array	The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available.	Yes
status	enum	Possible values: in_progress, completed, incomplete, interpreting, failed	Yes
type	enum	Possible values: code_interpreter_call	Yes

OpenAI.ComparisonFilter

A filter used to compare a specified attribute key to a given value using a defined comparison operation.

Name	Type	Description	Required	Default
key	string	The key to compare against the value.	Yes
type	enum	Specifies the comparison operator: eq, ne, gt, gte, lt, lte. - eq: equals - ne: not equal - gt: greater than - gte: greater than or equal - lt: less than - lte: less than or equal Possible values: eq, ne, gt, gte, lt, lte	Yes
value	string or number or boolean		Yes

OpenAI.CompletionUsage

Usage statistics for the completion request.

Name	Type	Description	Required	Default
completion_tokens	integer	Number of tokens in the generated completion.	Yes	0
completion_tokens_details	object	Breakdown of tokens used in a completion.	No
└─ accepted_prediction_tokens	integer	When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.	No	0
└─ audio_tokens	integer	Audio input tokens generated by the model.	No	0
└─ reasoning_tokens	integer	Tokens generated by the model for reasoning.	No	0
└─ rejected_prediction_tokens	integer	When using Predicted Outputs, the number of tokens in the prediction that did not appear in the completion. However, like reasoning tokens, these tokens are still counted in the total completion tokens for purposes of billing, output, and context window limits.	No	0
prompt_tokens	integer	Number of tokens in the prompt.	Yes	0
prompt_tokens_details	object	Breakdown of tokens used in the prompt.	No
└─ audio_tokens	integer	Audio input tokens present in the prompt.	No	0
└─ cached_tokens	integer	Cached tokens present in the prompt.	No	0
total_tokens	integer	Total number of tokens used in the request (prompt + completion).	Yes	0

OpenAI.CompoundFilter

Combine multiple filters using and or or.

Name	Type	Description	Required	Default
filters	array	Array of filters to combine. Items can be ComparisonFilter or CompoundFilter.	Yes
type	enum	Type of operation: and or or. Possible values: and, or	Yes

OpenAI.ComputerAction

Discriminator for OpenAI.ComputerAction

This component uses the property type to discriminate between different types:

Type Value	Schema
click	OpenAI.ComputerActionClick
double_click	OpenAI.ComputerActionDoubleClick
drag	OpenAI.ComputerActionDrag
move	OpenAI.ComputerActionMove
screenshot	OpenAI.ComputerActionScreenshot
scroll	OpenAI.ComputerActionScroll
type	OpenAI.ComputerActionTypeKeys
wait	OpenAI.ComputerActionWait
keypress	OpenAI.ComputerActionKeyPress

Name	Type	Description	Required	Default
type	OpenAI.ComputerActionType		Yes

OpenAI.ComputerActionClick

A click action.

Name	Type	Description	Required	Default
button	enum	Indicates which mouse button was pressed during the click. One of left, right, wheel, back, or forward. Possible values: left, right, wheel, back, forward	Yes
type	enum	Specifies the event type. For a click action, this property is always set to click. Possible values: click	Yes
x	integer	The x-coordinate where the click occurred.	Yes
y	integer	The y-coordinate where the click occurred.	Yes

OpenAI.ComputerActionDoubleClick

A double click action.

Name	Type	Description	Required	Default
type	enum	Specifies the event type. For a double click action, this property is always set to double_click. Possible values: double_click	Yes
x	integer	The x-coordinate where the double click occurred.	Yes
y	integer	The y-coordinate where the double click occurred.	Yes

OpenAI.ComputerActionDrag

A drag action.

Name	Type	Description	Required	Default
path	array	An array of coordinates representing the path of the drag action. Coordinates will appear as an array of objects, eg <br />[<br /> { x: 100, y: 200 },<br /> { x: 200, y: 300 }<br />]<br />	Yes
type	enum	Specifies the event type. For a drag action, this property is always set to drag. Possible values: drag	Yes

OpenAI.ComputerActionKeyPress

A collection of keypresses the model would like to perform.

Name	Type	Description	Required	Default
keys	array	The combination of keys the model is requesting to be pressed. This is an array of strings, each representing a key.	Yes
type	enum	Specifies the event type. For a keypress action, this property is always set to keypress. Possible values: keypress	Yes

OpenAI.ComputerActionMove

A mouse move action.

Name	Type	Description	Required	Default
type	enum	Specifies the event type. For a move action, this property is always set to move. Possible values: move	Yes
x	integer	The x-coordinate to move to.	Yes
y	integer	The y-coordinate to move to.	Yes

OpenAI.ComputerActionScreenshot

A screenshot action.

Name	Type	Description	Required	Default
type	enum	Specifies the event type. For a screenshot action, this property is always set to screenshot. Possible values: screenshot	Yes

OpenAI.ComputerActionScroll

A scroll action.

Name	Type	Description	Required	Default
scroll_x	integer	The horizontal scroll distance.	Yes
scroll_y	integer	The vertical scroll distance.	Yes
type	enum	Specifies the event type. For a scroll action, this property is always set to scroll. Possible values: scroll	Yes
x	integer	The x-coordinate where the scroll occurred.	Yes
y	integer	The y-coordinate where the scroll occurred.	Yes

OpenAI.ComputerActionType

Property	Value
Type	string
Values	screenshot click double_click scroll type wait keypress drag move

OpenAI.ComputerActionTypeKeys

An action to type in text.

Name	Type	Description	Required	Default
text	string	The text to type.	Yes
type	enum	Specifies the event type. For a type action, this property is always set to type. Possible values: type	Yes

OpenAI.ComputerActionWait

A wait action.

Name	Type	Description	Required	Default
type	enum	Specifies the event type. For a wait action, this property is always set to wait. Possible values: wait	Yes

OpenAI.ComputerToolCallItemParam

A tool call to a computer use tool.

Name	Type	Description	Required	Default
action	OpenAI.ComputerAction		Yes
call_id	string	An identifier used when responding to the tool call with output.	Yes
pending_safety_checks	array	The pending safety checks for the computer call.	Yes
type	enum	Possible values: computer_call	Yes

OpenAI.ComputerToolCallItemResource

A tool call to a computer use tool.

Name	Type	Description	Required	Default
action	OpenAI.ComputerAction		Yes
call_id	string	An identifier used when responding to the tool call with output.	Yes
pending_safety_checks	array	The pending safety checks for the computer call.	Yes
status	enum	The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API. Possible values: in_progress, completed, incomplete	Yes
type	enum	Possible values: computer_call	Yes

OpenAI.ComputerToolCallOutputItemOutput

Discriminator for OpenAI.ComputerToolCallOutputItemOutput

This component uses the property type to discriminate between different types:

Type Value	Schema
computer_screenshot	OpenAI.ComputerToolCallOutputItemOutputComputerScreenshot

Name	Type	Description	Required	Default
type	OpenAI.ComputerToolCallOutputItemOutputType	A computer screenshot image used with the computer use tool.	Yes

OpenAI.ComputerToolCallOutputItemOutputComputerScreenshot

Name	Type	Description	Required	Default
file_id	string		No
image_url	string		No
type	enum	Possible values: computer_screenshot	Yes

OpenAI.ComputerToolCallOutputItemOutputType

A computer screenshot image used with the computer use tool.

Property	Value
Description	A computer screenshot image used with the computer use tool.
Type	string
Values	computer_screenshot

OpenAI.ComputerToolCallOutputItemParam

The output of a computer tool call.

Name	Type	Description	Required	Default
acknowledged_safety_checks	array	The safety checks reported by the API that have been acknowledged by the developer.	No
call_id	string	The ID of the computer tool call that produced the output.	Yes
output	OpenAI.ComputerToolCallOutputItemOutput		Yes
type	enum	Possible values: computer_call_output	Yes

OpenAI.ComputerToolCallOutputItemResource

The output of a computer tool call.

Name	Type	Description	Required	Default
acknowledged_safety_checks	array	The safety checks reported by the API that have been acknowledged by the developer.	No
call_id	string	The ID of the computer tool call that produced the output.	Yes
output	OpenAI.ComputerToolCallOutputItemOutput		Yes
status	enum	The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API. Possible values: in_progress, completed, incomplete	Yes
type	enum	Possible values: computer_call_output	Yes

OpenAI.ComputerToolCallSafetyCheck

A pending safety check for the computer call.

Name	Type	Description	Required	Default
code	string	The type of the pending safety check.	Yes
id	string	The ID of the pending safety check.	Yes
message	string	Details about the pending safety check.	Yes

OpenAI.ComputerUsePreviewTool

A tool that controls a virtual computer.

Name	Type	Description	Required	Default
display_height	integer	The height of the computer display.	Yes
display_width	integer	The width of the computer display.	Yes
environment	enum	The type of computer environment to control. Possible values: windows, mac, linux, ubuntu, browser	Yes
type	enum	The type of the computer use tool. Always computer_use_preview. Possible values: computer_use_preview	Yes

OpenAI.Coordinate

An x/y coordinate pair, e.g. { x: 100, y: 200 }.

Name	Type	Description	Required	Default
x	integer	The x-coordinate.	Yes
y	integer	The y-coordinate.	Yes

OpenAI.CreateEmbeddingResponse

Name	Type	Description	Required	Default
data	array	The list of embeddings generated by the model.	Yes
model	string	The name of the model used to generate the embedding.	Yes
object	enum	The object type, which is always “list”. Possible values: list	Yes
usage	object	The usage information for the request.	Yes
└─ prompt_tokens	integer	The number of tokens used by the prompt.	No
└─ total_tokens	integer	The total number of tokens used by the request.	No

OpenAI.CreateEvalItem

A chat message that makes up the prompt or context. May include variable references to the item namespace, ie {{item.name}}.

Name	Type	Description	Required	Default
content	string or OpenAI.EvalItemContent	Text inputs to the model - can contain template strings.	Yes
role	enum	The role of the message input. One of user, assistant, system, or developer. Possible values: user, assistant, system, developer	Yes
type	enum	The type of the message input. Always message. Possible values: message	No

OpenAI.CreateEvalRunRequest

Name	Type	Description	Required	Default
data_source	object		Yes
└─ type	OpenAI.EvalRunDataSourceType		No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the run.	No

OpenAI.CreateFineTuningJobRequest

Valid models:

babbage-002
davinci-002
gpt-3.5-turbo
gpt-4o-mini

Name	Type	Description	Required	Default
hyperparameters	object	The hyperparameters used for the fine-tuning job. This value is now deprecated in favor of method, and should be passed in under the method parameter.	No
└─ batch_size	enum	Possible values: auto	No
└─ learning_rate_multiplier	enum	Possible values: auto	No
└─ n_epochs	enum	Possible values: auto	No
integrations	array	A list of integrations to enable for your fine-tuning job.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
method	OpenAI.FineTuneMethod	The method used for fine-tuning.	No
model	string (see valid models below)	The name of the model to fine-tune. You can select one of the supported models.	Yes
seed	integer	The seed controls the reproducibility of the job. Passing in the same seed and job parameters should produce the same results, but may differ in rare cases. If a seed is not specified, one will be generated for you.	No
suffix	string	A string of up to 64 characters that will be added to your fine-tuned model name. For example, a suffix of “custom-model-name” would produce a model name like ft:gpt-4o-mini:openai:custom-model-name:7p4lURel.	No	None
training_file	string	The ID of an uploaded file that contains training data. See upload file for how to upload a file. Your dataset must be formatted as a JSONL file. Additionally, you must upload your file with the purpose fine-tune. The contents of the file should differ depending on if the model uses the chat, or if the fine-tuning method uses the preference format. See the fine-tuning guide for more details.	Yes
validation_file	string	The ID of an uploaded file that contains validation data. If you provide this file, the data is used to generate validation metrics periodically during fine-tuning. These metrics can be viewed in the fine-tuning results file. The same data should not be present in both train and validation files. Your dataset must be formatted as a JSONL file. You must upload your file with the purpose fine-tune. See the fine-tuning guide for more details.	No

OpenAI.CreateFineTuningJobRequestIntegration

Discriminator for OpenAI.CreateFineTuningJobRequestIntegration

This component uses the property type to discriminate between different types:

Type Value	Schema
wandb	OpenAI.CreateFineTuningJobRequestWandbIntegration

Name	Type	Description	Required	Default
type	string (see valid models below)		Yes

OpenAI.CreateFineTuningJobRequestWandbIntegration

Name	Type	Description	Required	Default
type	enum	Possible values: wandb	Yes
wandb	object		Yes
└─ entity	string		No
└─ name	string		No
└─ project	string		No
└─ tags	array		No

OpenAI.CreateVectorStoreFileBatchRequest

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	No
chunking_strategy	OpenAI.ChunkingStrategyRequestParam	The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.	No
file_ids	array	A list of File IDs that the vector store should use. Useful for tools like file_search that can access files.	Yes

OpenAI.CreateVectorStoreFileRequest

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	No
chunking_strategy	OpenAI.ChunkingStrategyRequestParam	The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.	No
file_id	string	A File ID that the vector store should use. Useful for tools like file_search that can access files.	Yes

OpenAI.CreateVectorStoreRequest

Name	Type	Description	Required	Default
chunking_strategy	object	The default strategy. This strategy currently uses a max_chunk_size_tokens of 800 and chunk_overlap_tokens of 400.	No
└─ static	OpenAI.StaticChunkingStrategy		No
└─ type	enum	Always static. Possible values: static	No
expires_after	OpenAI.VectorStoreExpirationAfter	The expiration policy for a vector store.	No
file_ids	array	A list of File IDs that the vector store should use. Useful for tools like file_search that can access files.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the vector store.	No

OpenAI.DeleteFileResponse

Name	Type	Description	Required	Default
deleted	boolean		Yes
id	string		Yes
object	enum	Possible values: file	Yes

OpenAI.DeleteVectorStoreFileResponse

Name	Type	Description	Required	Default
deleted	boolean		Yes
id	string		Yes
object	enum	Possible values: vector_store.file.deleted	Yes

OpenAI.DeleteVectorStoreResponse

Name	Type	Description	Required	Default
deleted	boolean		Yes
id	string		Yes
object	enum	Possible values: vector_store.deleted	Yes

OpenAI.Embedding

Represents an embedding vector returned by embedding endpoint.

Name	Type	Description	Required	Default
embedding	array or string		Yes
index	integer	The index of the embedding in the list of embeddings.	Yes
object	enum	The object type, which is always “embedding”. Possible values: embedding	Yes

OpenAI.Eval

An Eval object with a data source config and testing criteria. An Eval represents a task to be done for your LLM integration. Like:

• Improve the quality of my chatbot
• See how well my chatbot handles customer support
• Check if o4-mini is better at my usecase than gpt-4o

Name	Type	Description	Required	Default
created_at	integer	The Unix timestamp (in seconds) for when the eval was created.	Yes
data_source_config	object		Yes
└─ type	OpenAI.EvalDataSourceConfigType		No
id	string	Unique identifier for the evaluation.	Yes
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
name	string	The name of the evaluation.	Yes
object	enum	The object type. Possible values: eval	Yes
testing_criteria	array	A list of testing criteria.	Yes	None

OpenAI.EvalApiError

An object representing an error response from the Eval API.

Name	Type	Description	Required	Default
code	string	The error code.	Yes
message	string	The error message.	Yes

OpenAI.EvalCompletionsRunDataSourceParams

A CompletionsRunDataSource object describing a model sampling configuration.

Name	Type	Description	Required	Default
input_messages	object		No
└─ item_reference	string	A reference to a variable in the item namespace. Ie, “item.input_trajectory”	No
└─ template	array	A list of chat messages forming the prompt or context. May include variable references to the item namespace, ie {{item.name}}.	No
└─ type	enum	The type of input messages. Always item_reference. Possible values: item_reference	No
model	string	The name of the model to use for generating completions (e.g. “o3-mini”).	No
sampling_params	AzureEvalAPICompletionsSamplingParams		No
source	object		Yes
└─ content	array	The content of the jsonl file.	No
└─ created_after	integer	An optional Unix timestamp to filter items created after this time.	No
└─ created_before	integer	An optional Unix timestamp to filter items created before this time.	No
└─ id	string	The identifier of the file.	No
└─ limit	integer	An optional maximum number of items to return.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ model	string	An optional model to filter by (e.g., ‘gpt-4o’).	No
└─ type	enum	The type of source. Always stored_completions. Possible values: stored_completions	No
type	enum	The type of run data source. Always completions. Possible values: completions	Yes

OpenAI.EvalCustomDataSourceConfigParams

A CustomDataSourceConfig object that defines the schema for the data source used for the evaluation runs. This schema is used to define the shape of the data that will be:

• Used to define your testing criteria and
• What data is required when creating a run

Name	Type	Description	Required	Default
include_sample_schema	boolean	Whether the eval should expect you to populate the sample namespace (ie, by generating responses off of your data source)	No	False
item_schema	object	The json schema for each row in the data source.	Yes
type	enum	The type of data source. Always custom. Possible values: custom	Yes

OpenAI.EvalCustomDataSourceConfigResource

A CustomDataSourceConfig which specifies the schema of your item and optionally sample namespaces. The response schema defines the shape of the data that will be:

• Used to define your testing criteria and
• What data is required when creating a run

Name	Type	Description	Required	Default
schema	object	The json schema for the run data source items. Learn how to build JSON schemas here.	Yes
type	enum	The type of data source. Always custom. Possible values: custom	Yes

OpenAI.EvalDataSourceConfigParams

Discriminator for OpenAI.EvalDataSourceConfigParams

This component uses the property type to discriminate between different types:

Type Value	Schema
custom	OpenAI.EvalCustomDataSourceConfigParams
logs	OpenAI.EvalLogsDataSourceConfigParams
stored_completions	OpenAI.EvalStoredCompletionsDataSourceConfigParams

Name	Type	Description	Required	Default
type	OpenAI.EvalDataSourceConfigType		Yes

OpenAI.EvalDataSourceConfigResource

Discriminator for OpenAI.EvalDataSourceConfigResource

This component uses the property type to discriminate between different types:

Type Value	Schema
custom	OpenAI.EvalCustomDataSourceConfigResource
stored_completions	OpenAI.EvalStoredCompletionsDataSourceConfigResource
logs	OpenAI.EvalLogsDataSourceConfigResource

Name	Type	Description	Required	Default
type	OpenAI.EvalDataSourceConfigType		Yes

OpenAI.EvalDataSourceConfigType

Property	Value
Type	string
Values	custom logs stored_completions

OpenAI.EvalGraderLabelModelParams

A LabelModelGrader object which uses a model to assign labels to each item in the evaluation.

Name	Type	Description	Required	Default
input	array	A list of chat messages forming the prompt or context. May include variable references to the item namespace, ie {{item.name}}.	Yes
labels	array	The labels to classify to each item in the evaluation.	Yes
model	string	The model to use for the evaluation. Must support structured outputs.	Yes
name	string	The name of the grader.	Yes
passing_labels	array	The labels that indicate a passing result. Must be a subset of labels.	Yes
type	enum	The object type, which is always label_model. Possible values: label_model	Yes

OpenAI.EvalGraderLabelModelResource

Name	Type	Description	Required	Default
input	array		Yes
labels	array	The labels to assign to each item in the evaluation.	Yes
model	string	The model to use for the evaluation. Must support structured outputs.	Yes
name	string	The name of the grader.	Yes
passing_labels	array	The labels that indicate a passing result. Must be a subset of labels.	Yes
type	enum	The object type, which is always label_model. Possible values: label_model	Yes

OpenAI.EvalGraderParams

Discriminator for OpenAI.EvalGraderParams

This component uses the property type to discriminate between different types:

Type Value	Schema
label_model	OpenAI.EvalGraderLabelModelParams
string_check	OpenAI.EvalGraderStringCheckParams
text_similarity	OpenAI.EvalGraderTextSimilarityParams
python	OpenAI.EvalGraderPythonParams
score_model	OpenAI.EvalGraderScoreModelParams

Name	Type	Description	Required	Default
type	OpenAI.GraderType		Yes

OpenAI.EvalGraderPythonParams

Name	Type	Description	Required	Default
image_tag	string	The image tag to use for the python script.	No
name	string	The name of the grader.	Yes
pass_threshold	number	The threshold for the score.	No
source	string	The source code of the python script.	Yes
type	enum	The object type, which is always python. Possible values: python	Yes

OpenAI.EvalGraderPythonResource

Name	Type	Description	Required	Default
image_tag	string	The image tag to use for the python script.	No
name	string	The name of the grader.	Yes
pass_threshold	number	The threshold for the score.	No
source	string	The source code of the python script.	Yes
type	enum	The object type, which is always python. Possible values: python	Yes

OpenAI.EvalGraderResource

Discriminator for OpenAI.EvalGraderResource

This component uses the property type to discriminate between different types:

Type Value	Schema
label_model	OpenAI.EvalGraderLabelModelResource
text_similarity	OpenAI.EvalGraderTextSimilarityResource
python	OpenAI.EvalGraderPythonResource
score_model	OpenAI.EvalGraderScoreModelResource

Name	Type	Description	Required	Default
type	OpenAI.GraderType		Yes

OpenAI.EvalGraderScoreModelParams

Name	Type	Description	Required	Default
input	array	The input text. This may include template strings.	Yes
model	string	The model to use for the evaluation.	Yes
name	string	The name of the grader.	Yes
pass_threshold	number	The threshold for the score.	No
range	array	The range of the score. Defaults to [0, 1].	No
sampling_params		The sampling parameters for the model.	No
type	enum	The object type, which is always score_model. Possible values: score_model	Yes

OpenAI.EvalGraderScoreModelResource

Name	Type	Description	Required	Default
input	array	The input text. This may include template strings.	Yes
model	string	The model to use for the evaluation.	Yes
name	string	The name of the grader.	Yes
pass_threshold	number	The threshold for the score.	No
range	array	The range of the score. Defaults to [0, 1].	No
sampling_params		The sampling parameters for the model.	No
type	enum	The object type, which is always score_model. Possible values: score_model	Yes

OpenAI.EvalGraderStringCheckParams

Name	Type	Description	Required	Default
input	string	The input text. This may include template strings.	Yes
name	string	The name of the grader.	Yes
operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	Yes
reference	string	The reference text. This may include template strings.	Yes
type	enum	The object type, which is always string_check. Possible values: string_check	Yes

OpenAI.EvalGraderTextSimilarityParams

Name	Type	Description	Required	Default
evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	Yes
input	string	The text being graded.	Yes
name	string	The name of the grader.	Yes
pass_threshold	number	The threshold for the score.	Yes
reference	string	The text being graded against.	Yes
type	enum	The type of grader. Possible values: text_similarity	Yes

OpenAI.EvalGraderTextSimilarityResource

Name	Type	Description	Required	Default
evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	Yes
input	string	The text being graded.	Yes
name	string	The name of the grader.	Yes
pass_threshold	number	The threshold for the score.	Yes
reference	string	The text being graded against.	Yes
type	enum	The type of grader. Possible values: text_similarity	Yes

OpenAI.EvalItem

A message input to the model with a role indicating instruction following hierarchy. Instructions given with the developer or system role take precedence over instructions given with the user role. Messages with the assistant role are presumed to have been generated by the model in previous interactions.

Name	Type	Description	Required	Default
content	object		Yes
└─ type	OpenAI.EvalItemContentType		No
role	enum	The role of the message input. One of user, assistant, system, or developer. Possible values: user, assistant, system, developer	Yes
type	enum	The type of the message input. Always message. Possible values: message	No

OpenAI.EvalItemContent

Discriminator for OpenAI.EvalItemContent

This component uses the property type to discriminate between different types:

Type Value	Schema
input_text	OpenAI.EvalItemContentInputText
output_text	OpenAI.EvalItemContentOutputText

Name	Type	Description	Required	Default
type	OpenAI.EvalItemContentType		Yes

OpenAI.EvalItemContentInputText

Name	Type	Description	Required	Default
text	string		Yes
type	enum	Possible values: input_text	Yes

OpenAI.EvalItemContentOutputText

Name	Type	Description	Required	Default
text	string		Yes
type	enum	Possible values: output_text	Yes

OpenAI.EvalItemContentType

Property	Value
Type	string
Values	input_text output_text

OpenAI.EvalJsonlRunDataSourceParams

A JsonlRunDataSource object with that specifies a JSONL file that matches the eval

Name	Type	Description	Required	Default
source	object		Yes
└─ content	array	The content of the jsonl file.	No
└─ id	string	The identifier of the file.	No
└─ type	enum	The type of jsonl source. Always file_id. Possible values: file_id	No
type	enum	The type of data source. Always jsonl. Possible values: jsonl	Yes

OpenAI.EvalList

An object representing a list of evals.

Name	Type	Description	Required	Default
data	array	An array of eval objects.	Yes
first_id	string	The identifier of the first eval in the data array.	Yes
has_more	boolean	Indicates whether there are more evals available.	Yes
last_id	string	The identifier of the last eval in the data array.	Yes
object	enum	The type of this object. It is always set to “list”. Possible values: list	Yes

OpenAI.EvalLogsDataSourceConfigParams

A data source config which specifies the metadata property of your logs query. This is usually metadata like usecase=chatbot or prompt-version=v2, etc.

Name	Type	Description	Required	Default
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
type	enum	The type of data source. Always logs. Possible values: logs	Yes

OpenAI.EvalLogsDataSourceConfigResource

A LogsDataSourceConfig which specifies the metadata property of your logs query. This is usually metadata like usecase=chatbot or prompt-version=v2, etc. The schema returned by this data source config is used to defined what variables are available in your evals. item and sample are both defined when using this data source config.

Name	Type	Description	Required	Default
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
schema	object	The json schema for the run data source items. Learn how to build JSON schemas here.	Yes
type	enum	The type of data source. Always logs. Possible values: logs	Yes

OpenAI.EvalResponsesRunDataSourceParams

A ResponsesRunDataSource object describing a model sampling configuration.

Name	Type	Description	Required	Default
input_messages	object		No
└─ item_reference	string	A reference to a variable in the item namespace. Ie, “item.name”	No
└─ template	array	A list of chat messages forming the prompt or context. May include variable references to the item namespace, ie {{item.name}}.	No
└─ type	enum	The type of input messages. Always item_reference. Possible values: item_reference	No
model	string	The name of the model to use for generating completions (e.g. “o3-mini”).	No
sampling_params	AzureEvalAPIResponseSamplingParams		No
source	object		Yes
└─ content	array	The content of the jsonl file.	No
└─ created_after	integer	Only include items created after this timestamp (inclusive). This is a query parameter used to select responses.	No
└─ created_before	integer	Only include items created before this timestamp (inclusive). This is a query parameter used to select responses.	No
└─ id	string	The identifier of the file.	No
└─ instructions_search	string	Optional string to search the ‘instructions’ field. This is a query parameter used to select responses.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ model	string	The name of the model to find responses for. This is a query parameter used to select responses.	No
└─ reasoning_effort	OpenAI.ReasoningEffort	Optional reasoning effort parameter. This is a query parameter used to select responses.	No
└─ temperature	number	Sampling temperature. This is a query parameter used to select responses.	No
└─ tools	array	List of tool names. This is a query parameter used to select responses.	No
└─ top_p	number	Nucleus sampling parameter. This is a query parameter used to select responses.	No
└─ type	enum	The type of run data source. Always responses. Possible values: responses	No
└─ users	array	List of user identifiers. This is a query parameter used to select responses.	No
type	enum	The type of run data source. Always responses. Possible values: responses	Yes

OpenAI.EvalRun

A schema representing an evaluation run.

Name	Type	Description	Required	Default
created_at	integer	Unix timestamp (in seconds) when the evaluation run was created.	Yes
data_source	object		Yes
└─ type	OpenAI.EvalRunDataSourceType		No
error	OpenAI.EvalApiError	An object representing an error response from the Eval API.	Yes
eval_id	string	The identifier of the associated evaluation.	Yes
id	string	Unique identifier for the evaluation run.	Yes
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
model	string	The model that is evaluated, if applicable.	Yes
name	string	The name of the evaluation run.	Yes
object	enum	The type of the object. Always “eval.run”. Possible values: eval.run	Yes
per_model_usage	array	Usage statistics for each model during the evaluation run.	Yes
per_testing_criteria_results	array	Results per testing criteria applied during the evaluation run.	Yes
report_url	string	The URL to the rendered evaluation run report on the UI dashboard.	Yes
result_counts	object	Counters summarizing the outcomes of the evaluation run.	Yes
└─ errored	integer	Number of output items that resulted in an error.	No
└─ failed	integer	Number of output items that failed to pass the evaluation.	No
└─ passed	integer	Number of output items that passed the evaluation.	No
└─ total	integer	Total number of executed output items.	No
status	string	The status of the evaluation run.	Yes

OpenAI.EvalRunDataContentSource

Discriminator for OpenAI.EvalRunDataContentSource

This component uses the property type to discriminate between different types:

Type Value	Schema
file_id	OpenAI.EvalRunFileIdDataContentSource
stored_completions	OpenAI.EvalRunStoredCompletionsDataContentSource
responses	OpenAI.EvalRunResponsesDataContentSource

Name	Type	Description	Required	Default
type	OpenAI.EvalRunDataContentSourceType		Yes

OpenAI.EvalRunDataContentSourceType

Property	Value
Type	string
Values	file_id file_content stored_completions responses

OpenAI.EvalRunDataSourceCompletionsResource

Name	Type	Description	Required	Default
type	enum	Possible values: completions	Yes

OpenAI.EvalRunDataSourceJsonlResource

Name	Type	Description	Required	Default
type	enum	Possible values: jsonl	Yes

OpenAI.EvalRunDataSourceParams

Discriminator for OpenAI.EvalRunDataSourceParams

This component uses the property type to discriminate between different types:

Type Value	Schema
jsonl	OpenAI.EvalJsonlRunDataSourceParams
completions	OpenAI.EvalCompletionsRunDataSourceParams
responses	OpenAI.EvalResponsesRunDataSourceParams

Name	Type	Description	Required	Default
type	OpenAI.EvalRunDataSourceType		Yes

OpenAI.EvalRunDataSourceResource

Name	Type	Description	Required	Default
type	OpenAI.EvalRunDataSourceType		Yes

OpenAI.EvalRunDataSourceResponsesResource

Name	Type	Description	Required	Default
type	enum	Possible values: responses	Yes

OpenAI.EvalRunDataSourceType

Property	Value
Type	string
Values	jsonl completions responses

OpenAI.EvalRunFileContentDataContentSource

Name	Type	Description	Required	Default
content	array	The content of the jsonl file.	Yes
type	enum	The type of jsonl source. Always file_content. Possible values: file_content	Yes

OpenAI.EvalRunFileIdDataContentSource

Name	Type	Description	Required	Default
id	string	The identifier of the file.	Yes
type	enum	The type of jsonl source. Always file_id. Possible values: file_id	Yes

OpenAI.EvalRunList

An object representing a list of runs for an evaluation.

Name	Type	Description	Required	Default
data	array	An array of eval run objects.	Yes
first_id	string	The identifier of the first eval run in the data array.	Yes
has_more	boolean	Indicates whether there are more evals available.	Yes
last_id	string	The identifier of the last eval run in the data array.	Yes
object	enum	The type of this object. It is always set to “list”. Possible values: list	Yes

OpenAI.EvalRunOutputItem

A schema representing an evaluation run output item.

Name	Type	Description	Required	Default
created_at	integer	Unix timestamp (in seconds) when the evaluation run was created.	Yes
datasource_item	object	Details of the input data source item.	Yes
datasource_item_id	integer	The identifier for the data source item.	Yes
eval_id	string	The identifier of the evaluation group.	Yes
id	string	Unique identifier for the evaluation run output item.	Yes
object	enum	The type of the object. Always “eval.run.output_item”. Possible values: eval.run.output_item	Yes
results	array	A list of results from the evaluation run.	Yes
run_id	string	The identifier of the evaluation run associated with this output item.	Yes
sample	object	A sample containing the input and output of the evaluation run.	Yes
└─ error	OpenAI.EvalApiError	An object representing an error response from the Eval API.	No
└─ finish_reason	string	The reason why the sample generation was finished.	No
└─ input	array	An array of input messages.	No
└─ max_completion_tokens	integer	The maximum number of tokens allowed for completion.	No
└─ model	string	The model used for generating the sample.	No
└─ output	array	An array of output messages.	No
└─ seed	integer	The seed used for generating the sample.	No
└─ temperature	number	The sampling temperature used.	No
└─ top_p	number	The top_p value used for sampling.	No
└─ usage	object	Token usage details for the sample.	No
└─ cached_tokens	integer	The number of tokens retrieved from cache.	No
└─ completion_tokens	integer	The number of completion tokens generated.	No
└─ prompt_tokens	integer	The number of prompt tokens used.	No
└─ total_tokens	integer	The total number of tokens used.	No
status	string	The status of the evaluation run.	Yes

OpenAI.EvalRunOutputItemList

An object representing a list of output items for an evaluation run.

Name	Type	Description	Required	Default
data	array	An array of eval run output item objects.	Yes
first_id	string	The identifier of the first eval run output item in the data array.	Yes
has_more	boolean	Indicates whether there are more eval run output items available.	Yes
last_id	string	The identifier of the last eval run output item in the data array.	Yes
object	enum	The type of this object. It is always set to “list”. Possible values: list	Yes

OpenAI.EvalRunResponsesDataContentSource

A EvalResponsesSource object describing a run data source configuration.

Name	Type	Description	Required	Default
created_after	integer	Only include items created after this timestamp (inclusive). This is a query parameter used to select responses.	No
created_before	integer	Only include items created before this timestamp (inclusive). This is a query parameter used to select responses.	No
instructions_search	string	Optional string to search the ‘instructions’ field. This is a query parameter used to select responses.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
model	string	The name of the model to find responses for. This is a query parameter used to select responses.	No
reasoning_effort	object	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
temperature	number	Sampling temperature. This is a query parameter used to select responses.	No
tools	array	List of tool names. This is a query parameter used to select responses.	No
top_p	number	Nucleus sampling parameter. This is a query parameter used to select responses.	No
type	enum	The type of run data source. Always responses. Possible values: responses	Yes
users	array	List of user identifiers. This is a query parameter used to select responses.	No

OpenAI.EvalRunStoredCompletionsDataContentSource

A StoredCompletionsRunDataSource configuration describing a set of filters

Name	Type	Description	Required	Default
created_after	integer	An optional Unix timestamp to filter items created after this time.	No
created_before	integer	An optional Unix timestamp to filter items created before this time.	No
limit	integer	An optional maximum number of items to return.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
model	string	An optional model to filter by (e.g., ‘gpt-4o’).	No
type	enum	The type of source. Always stored_completions. Possible values: stored_completions	Yes

OpenAI.EvalStoredCompletionsDataSourceConfigParams

Deprecated in favor of LogsDataSourceConfig.

Name	Type	Description	Required	Default
metadata	object	Metadata filters for the stored completions data source.	No
type	enum	The type of data source. Always stored_completions. Possible values: stored_completions	Yes

OpenAI.EvalStoredCompletionsDataSourceConfigResource

Deprecated in favor of LogsDataSourceConfig.

Name	Type	Description	Required	Default
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
schema	object	The json schema for the run data source items. Learn how to build JSON schemas here.	Yes
type	enum	The type of data source. Always stored_completions. Possible values: stored_completions	Yes

OpenAI.FileSearchTool

A tool that searches for relevant content from uploaded files.

Name	Type	Description	Required	Default
filters	object		No
max_num_results	integer	The maximum number of results to return. This number should be between 1 and 50 inclusive.	No
ranking_options	object		No
└─ ranker	enum	The ranker to use for the file search. Possible values: auto, default-2024-11-15	No
└─ score_threshold	number	The score threshold for the file search, a number between 0 and 1. Numbers closer to 1 will attempt to return only the most relevant results, but may return fewer results.	No
type	enum	The type of the file search tool. Always file_search. Possible values: file_search	Yes
vector_store_ids	array	The IDs of the vector stores to search.	Yes

OpenAI.FileSearchToolCallItemParam

The results of a file search tool call.

Name	Type	Description	Required	Default
queries	array	The queries used to search for files.	Yes
results	array	The results of the file search tool call.	No
type	enum	Possible values: file_search_call	Yes

OpenAI.FileSearchToolCallItemResource

The results of a file search tool call.

Name	Type	Description	Required	Default
queries	array	The queries used to search for files.	Yes
results	array	The results of the file search tool call.	No
status	enum	The status of the file search tool call. One of in_progress, searching, incomplete or failed, Possible values: in_progress, searching, completed, incomplete, failed	Yes
type	enum	Possible values: file_search_call	Yes

OpenAI.Filters

Name	Type	Description	Required	Default
filters	array	Array of filters to combine. Items can be ComparisonFilter or CompoundFilter.	Yes
key	string	The key to compare against the value.	Yes
type	enum	Type of operation: and or or. Possible values: and, or	Yes
value	string or number or boolean	The value to compare against the attribute key; supports string, number, or boolean types.	Yes

OpenAI.FineTuneDPOHyperparameters

The hyperparameters used for the DPO fine-tuning job.

Name	Type	Description	Required	Default
batch_size	enum	Possible values: auto	No
beta	enum	Possible values: auto	No
learning_rate_multiplier	enum	Possible values: auto	No
n_epochs	enum	Possible values: auto	No

OpenAI.FineTuneDPOMethod

Configuration for the DPO fine-tuning method.

Name	Type	Description	Required	Default
hyperparameters	OpenAI.FineTuneDPOHyperparameters	The hyperparameters used for the DPO fine-tuning job.	No

OpenAI.FineTuneMethod

The method used for fine-tuning.

Name	Type	Description	Required	Default
dpo	OpenAI.FineTuneDPOMethod	Configuration for the DPO fine-tuning method.	No
reinforcement	AzureFineTuneReinforcementMethod		No
supervised	OpenAI.FineTuneSupervisedMethod	Configuration for the supervised fine-tuning method.	No
type	enum	The type of method. Is either supervised, dpo, or reinforcement. Possible values: supervised, dpo, reinforcement	Yes

OpenAI.FineTuneReinforcementHyperparameters

The hyperparameters used for the reinforcement fine-tuning job.

Name	Type	Description	Required	Default
batch_size	enum	Possible values: auto	No
compute_multiplier	enum	Possible values: auto	No
eval_interval	enum	Possible values: auto	No
eval_samples	enum	Possible values: auto	No
learning_rate_multiplier	enum	Possible values: auto	No
n_epochs	enum	Possible values: auto	No
reasoning_effort	enum	Level of reasoning effort. Possible values: default, low, medium, high	No

OpenAI.FineTuneSupervisedHyperparameters

The hyperparameters used for the fine-tuning job.

Name	Type	Description	Required	Default
batch_size	enum	Possible values: auto	No
learning_rate_multiplier	enum	Possible values: auto	No
n_epochs	enum	Possible values: auto	No

OpenAI.FineTuneSupervisedMethod

Configuration for the supervised fine-tuning method.

Name	Type	Description	Required	Default
hyperparameters	OpenAI.FineTuneSupervisedHyperparameters	The hyperparameters used for the fine-tuning job.	No

OpenAI.FineTuningIntegration

Discriminator for OpenAI.FineTuningIntegration

This component uses the property type to discriminate between different types:

Type Value	Schema
wandb	OpenAI.FineTuningIntegrationWandb

Name	Type	Description	Required	Default
type	string (see valid models below)		Yes

OpenAI.FineTuningIntegrationWandb

Name	Type	Description	Required	Default
type	enum	The type of the integration being enabled for the fine-tuning job Possible values: wandb	Yes
wandb	object	The settings for your integration with Weights and Biases. This payload specifies the project that metrics will be sent to. Optionally, you can set an explicit display name for your run, add tags to your run, and set a default entity (team, username, etc) to be associated with your run.	Yes
└─ entity	string	The entity to use for the run. This allows you to set the team or username of the WandB user that you would like associated with the run. If not set, the default entity for the registered WandB API key is used.	No
└─ name	string	A display name to set for the run. If not set, we will use the Job ID as the name.	No
└─ project	string	The name of the project that the new run will be created under.	No
└─ tags	array	A list of tags to be attached to the newly created run. These tags are passed through directly to WandB. Some default tags are generated by OpenAI: “openai/finetune”, “openai/{base-model}”, “openai/{ftjob-abcdef}”.	No

OpenAI.FineTuningJob

The fine_tuning.job object represents a fine-tuning job that has been created through the API.

Name	Type	Description	Required	Default
created_at	integer	The Unix timestamp (in seconds) for when the fine-tuning job was created.	Yes
error	object	For fine-tuning jobs that have failed, this will contain more information on the cause of the failure.	Yes
└─ code	string	A machine-readable error code.	No
└─ message	string	A human-readable error message.	No
└─ param	string	The parameter that was invalid, usually training_file or validation_file. This field will be null if the failure was not parameter-specific.	No
estimated_finish	integer	The Unix timestamp (in seconds) for when the fine-tuning job is estimated to finish. The value will be null if the fine-tuning job is not running.	No
fine_tuned_model	string	The name of the fine-tuned model that is being created. The value will be null if the fine-tuning job is still running.	Yes
finished_at	integer	The Unix timestamp (in seconds) for when the fine-tuning job was finished. The value will be null if the fine-tuning job is still running.	Yes
hyperparameters	object	The hyperparameters used for the fine-tuning job. This value will only be returned when running supervised jobs.	Yes
└─ batch_size	enum	Possible values: auto	No
└─ learning_rate_multiplier	enum	Possible values: auto	No
└─ n_epochs	enum	Possible values: auto	No
id	string	The object identifier, which can be referenced in the API endpoints.	Yes
integrations	array	A list of integrations to enable for this fine-tuning job.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
method	OpenAI.FineTuneMethod	The method used for fine-tuning.	No
model	string	The base model that is being fine-tuned.	Yes
object	enum	The object type, which is always “fine_tuning.job”. Possible values: fine_tuning.job	Yes
organization_id	string	The organization that owns the fine-tuning job.	Yes
result_files	array	The compiled results file ID(s) for the fine-tuning job. You can retrieve the results with the Files API.	Yes
seed	integer	The seed used for the fine-tuning job.	Yes
status	enum	The current status of the fine-tuning job, which can be either validating_files, queued, running, succeeded, failed, or cancelled. Possible values: validating_files, queued, running, succeeded, failed, cancelled	Yes
trained_tokens	integer	The total number of billable tokens processed by this fine-tuning job. The value will be null if the fine-tuning job is still running.	Yes
training_file	string	The file ID used for training. You can retrieve the training data with the Files API.	Yes
user_provided_suffix	string	The descriptive suffix applied to the job, as specified in the job creation request.	No
validation_file	string	The file ID used for validation. You can retrieve the validation results with the Files API.	Yes

OpenAI.FineTuningJobCheckpoint

The fine_tuning.job.checkpoint object represents a model checkpoint for a fine-tuning job that is ready to use.

Name	Type	Description	Required	Default
created_at	integer	The Unix timestamp (in seconds) for when the checkpoint was created.	Yes
fine_tuned_model_checkpoint	string	The name of the fine-tuned checkpoint model that is created.	Yes
fine_tuning_job_id	string	The name of the fine-tuning job that this checkpoint was created from.	Yes
id	string	The checkpoint identifier, which can be referenced in the API endpoints.	Yes
metrics	object	Metrics at the step number during the fine-tuning job.	Yes
└─ full_valid_loss	number		No
└─ full_valid_mean_token_accuracy	number		No
└─ step	number		No
└─ train_loss	number		No
└─ train_mean_token_accuracy	number		No
└─ valid_loss	number		No
└─ valid_mean_token_accuracy	number		No
object	enum	The object type, which is always “fine_tuning.job.checkpoint”. Possible values: fine_tuning.job.checkpoint	Yes
step_number	integer	The step number that the checkpoint was created at.	Yes

OpenAI.FineTuningJobEvent

Fine-tuning job event object

Name	Type	Description	Required	Default
created_at	integer	The Unix timestamp (in seconds) for when the fine-tuning job was created.	Yes
data		The data associated with the event.	No
id	string	The object identifier.	Yes
level	enum	The log level of the event. Possible values: info, warn, error	Yes
message	string	The message of the event.	Yes
object	enum	The object type, which is always “fine_tuning.job.event”. Possible values: fine_tuning.job.event	Yes
type	enum	The type of event. Possible values: message, metrics	No

OpenAI.FunctionObject

Name	Type	Description	Required	Default
description	string	A description of what the function does, used by the model to choose when and how to call the function.	No
name	string	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.	Yes
parameters		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. Omitting parameters defines a function with an empty parameter list.	No
strict	boolean	Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.	No	False

OpenAI.FunctionTool

Defines a function in your own code the model can choose to call. Learn more about function calling.

Name	Type	Description	Required	Default
description	string	A description of the function. Used by the model to determine whether or not to call the function.	No
name	string	The name of the function to call.	Yes
parameters		A JSON schema object describing the parameters of the function.	Yes
strict	boolean	Whether to enforce strict parameter validation. Default true.	Yes
type	enum	The type of the function tool. Always function. Possible values: function	Yes

OpenAI.FunctionToolCallItemParam

A tool call to run a function. See the function calling guide for more information.

Name	Type	Description	Required	Default
arguments	string	A JSON string of the arguments to pass to the function.	Yes
call_id	string	The unique ID of the function tool call generated by the model.	Yes
name	string	The name of the function to run.	Yes
type	enum	Possible values: function_call	Yes

OpenAI.FunctionToolCallItemResource

A tool call to run a function. See the function calling guide for more information.

Name	Type	Description	Required	Default
arguments	string	A JSON string of the arguments to pass to the function.	Yes
call_id	string	The unique ID of the function tool call generated by the model.	Yes
name	string	The name of the function to run.	Yes
status	enum	The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API. Possible values: in_progress, completed, incomplete	Yes
type	enum	Possible values: function_call	Yes

OpenAI.FunctionToolCallOutputItemParam

The output of a function tool call.

Name	Type	Description	Required	Default
call_id	string	The unique ID of the function tool call generated by the model.	Yes
output	string	A JSON string of the output of the function tool call.	Yes
type	enum	Possible values: function_call_output	Yes

OpenAI.FunctionToolCallOutputItemResource

The output of a function tool call.

Name	Type	Description	Required	Default
call_id	string	The unique ID of the function tool call generated by the model.	Yes
output	string	A JSON string of the output of the function tool call.	Yes
status	enum	The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API. Possible values: in_progress, completed, incomplete	Yes
type	enum	Possible values: function_call_output	Yes

OpenAI.Grader

Discriminator for OpenAI.Grader

This component uses the property type to discriminate between different types:

Type Value	Schema
label_model	OpenAI.GraderLabelModel
text_similarity	OpenAI.GraderTextSimilarity
python	OpenAI.GraderPython
score_model	OpenAI.GraderScoreModel
multi	OpenAI.GraderMulti

Name	Type	Description	Required	Default
type	OpenAI.GraderType		Yes

OpenAI.GraderLabelModel

A LabelModelGrader object which uses a model to assign labels to each item in the evaluation.

Name	Type	Description	Required	Default
input	array		Yes
labels	array	The labels to assign to each item in the evaluation.	Yes
model	string	The model to use for the evaluation. Must support structured outputs.	Yes
name	string	The name of the grader.	Yes
passing_labels	array	The labels that indicate a passing result. Must be a subset of labels.	Yes
type	enum	The object type, which is always label_model. Possible values: label_model	Yes

OpenAI.GraderMulti

A MultiGrader object combines the output of multiple graders to produce a single score.

Name	Type	Description	Required	Default
calculate_output	string	A formula to calculate the output based on grader results.	Yes
graders	object		Yes
name	string	The name of the grader.	Yes
type	enum	The object type, which is always multi. Possible values: multi	Yes

OpenAI.GraderPython

A PythonGrader object that runs a python script on the input.

Name	Type	Description	Required	Default
image_tag	string	The image tag to use for the python script.	No
name	string	The name of the grader.	Yes
source	string	The source code of the python script.	Yes
type	enum	The object type, which is always python. Possible values: python	Yes

OpenAI.GraderScoreModel

A ScoreModelGrader object that uses a model to assign a score to the input.

Name	Type	Description	Required	Default
input	array	The input text. This may include template strings.	Yes
model	string	The model to use for the evaluation.	Yes
name	string	The name of the grader.	Yes
range	array	The range of the score. Defaults to [0, 1].	No
sampling_params		The sampling parameters for the model.	No
type	enum	The object type, which is always score_model. Possible values: score_model	Yes

OpenAI.GraderStringCheck

A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.

Name	Type	Description	Required	Default
input	string	The input text. This may include template strings.	Yes
name	string	The name of the grader.	Yes
operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	Yes
reference	string	The reference text. This may include template strings.	Yes
type	enum	The object type, which is always string_check. Possible values: string_check	Yes

OpenAI.GraderTextSimilarity

A TextSimilarityGrader object which grades text based on similarity metrics.

Name	Type	Description	Required	Default
evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	Yes
input	string	The text being graded.	Yes
name	string	The name of the grader.	Yes
reference	string	The text being graded against.	Yes
type	enum	The type of grader. Possible values: text_similarity	Yes

OpenAI.GraderType

Property	Value
Type	string
Values	string_check text_similarity score_model label_model python multi

OpenAI.ImageGenTool

A tool that generates images using a model like gpt-image-1-series.

Name	Type	Description	Required	Default
background	enum	Background type for the generated image. One of transparent, opaque, or auto. Default: auto. Possible values: transparent, opaque, auto	No
input_image_mask	object	Optional mask for inpainting. Contains image_url (string, optional) and file_id (string, optional).	No
└─ file_id	string	File ID for the mask image.	No
└─ image_url	string	Base64-encoded mask image.	No
model	enum	The image generation model to use. Default: gpt-image-1. Possible values: gpt-image-1	No
moderation	enum	Moderation level for the generated image. Default: auto. Possible values: auto, low	No
output_compression	integer	Compression level for the output image. Default: 100.	No	100
output_format	enum	The output format of the generated image. One of png, webp, or jpeg. Default: png. Possible values: png, webp, jpeg	No
partial_images	integer	Number of partial images to generate in streaming mode, from 0 (default value) to 3.	No	0
quality	enum	The quality of the generated image. One of low, medium, high, or auto. Default: auto. Possible values: low, medium, high, auto	No
size	enum	The size of the generated image. One of 1024x1024, 1024x1536, 1536x1024, or auto. Default: auto. Possible values: 1024x1024, 1024x1536, 1536x1024, auto	No
type	enum	The type of the image generation tool. Always image_generation. Possible values: image_generation	Yes

OpenAI.ImageGenToolCallItemParam

An image generation request made by the model.

Name	Type	Description	Required	Default
result	string	The generated image encoded in base64.	Yes
type	enum	Possible values: image_generation_call	Yes

OpenAI.ImageGenToolCallItemResource

An image generation request made by the model.

Name	Type	Description	Required	Default
result	string	The generated image encoded in base64.	Yes
status	enum	Possible values: in_progress, completed, generating, failed	Yes
type	enum	Possible values: image_generation_call	Yes

OpenAI.ImplicitUserMessage

Name	Type	Description	Required	Default
content	string or array		Yes

OpenAI.Includable

Specify additional output data to include in the model response. Currently supported values are:

• code_interpreter_call.outputs: Includes the outputs of python code execution in code interpreter tool call items.
• computer_call_output.output.image_url: Include image urls from the computer call output.
• file_search_call.results: Include the search results of the file search tool call.
• message.input_image.image_url: Include image urls from the input message.
• message.output_text.logprobs: Include logprobs with assistant messages.
• reasoning.encrypted_content: Includes an encrypted version of reasoning tokens in reasoning item outputs. This enables reasoning items to be used in multi-turn conversations when using the Responses API statelessly (like when the store parameter is set to false, or when an organization is enrolled in the zero data retention program).

Property	Value
Description	Specify additional output data to include in the model response. Currently supported values are: - code_interpreter_call.outputs: Includes the outputs of python code execution in code interpreter tool call items. - computer_call_output.output.image_url: Include image urls from the computer call output. - file_search_call.results: Include the search results of the file search tool call. - message.input_image.image_url: Include image urls from the input message. - message.output_text.logprobs: Include logprobs with assistant messages. - reasoning.encrypted_content: Includes an encrypted version of reasoning tokens in reasoning item outputs. This enables reasoning items to be used in multi-turn conversations when using the Responses API statelessly (like when the store parameter is set to false, or when an organization is enrolled in the zero data retention program).
Type	string
Values	code_interpreter_call.outputs computer_call_output.output.image_url file_search_call.results message.input_image.image_url message.output_text.logprobs reasoning.encrypted_content

OpenAI.ItemContent

Discriminator for OpenAI.ItemContent

This component uses the property type to discriminate between different types:

Type Value	Schema
input_audio	OpenAI.ItemContentInputAudio
output_audio	OpenAI.ItemContentOutputAudio
refusal	OpenAI.ItemContentRefusal
input_text	OpenAI.ItemContentInputText
input_image	OpenAI.ItemContentInputImage
input_file	OpenAI.ItemContentInputFile
output_text	OpenAI.ItemContentOutputText

Name	Type	Description	Required	Default
type	OpenAI.ItemContentType	Multi-modal input and output contents.	Yes

OpenAI.ItemContentInputAudio

An audio input to the model.

Name	Type	Description	Required	Default
data	string	Base64-encoded audio data.	Yes
format	enum	The format of the audio data. Currently supported formats are mp3 and wav. Possible values: mp3, wav	Yes
type	enum	The type of the input item. Always input_audio. Possible values: input_audio	Yes

OpenAI.ItemContentInputFile

A file input to the model.

Name	Type	Description	Required	Default
file_data	string	The content of the file to be sent to the model.	No
file_id	string	The ID of the file to be sent to the model.	No
filename	string	The name of the file to be sent to the model.	No
type	enum	The type of the input item. Always input_file. Possible values: input_file	Yes

OpenAI.ItemContentInputImage

An image input to the model.

Name	Type	Description	Required	Default
detail	enum	The detail level of the image to be sent to the model. One of high, low, or auto. Defaults to auto. Possible values: low, high, auto	No
file_id	string	The ID of the file to be sent to the model.	No
image_url	string	The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL.	No
type	enum	The type of the input item. Always input_image. Possible values: input_image	Yes

OpenAI.ItemContentInputText

A text input to the model.

Name	Type	Description	Required	Default
text	string	The text input to the model.	Yes
type	enum	The type of the input item. Always input_text. Possible values: input_text	Yes

OpenAI.ItemContentOutputAudio

An audio output from the model.

Name	Type	Description	Required	Default
data	string	Base64-encoded audio data from the model.	Yes
transcript	string	The transcript of the audio data from the model.	Yes
type	enum	The type of the output audio. Always output_audio. Possible values: output_audio	Yes

OpenAI.ItemContentOutputText

A text output from the model.

Name	Type	Description	Required	Default
annotations	array	The annotations of the text output.	Yes
logprobs	array		No
text	string	The text output from the model.	Yes
type	enum	The type of the output text. Always output_text. Possible values: output_text	Yes

OpenAI.ItemContentRefusal

A refusal from the model.

Name	Type	Description	Required	Default
refusal	string	The refusal explanationfrom the model.	Yes
type	enum	The type of the refusal. Always refusal. Possible values: refusal	Yes

OpenAI.ItemContentType

Multi-modal input and output contents.

Property	Value
Description	Multi-modal input and output contents.
Type	string
Values	input_text input_audio input_image input_file output_text output_audio refusal

OpenAI.ItemParam

Content item used to generate a response.

Discriminator for OpenAI.ItemParam

This component uses the property type to discriminate between different types:

Type Value	Schema
file_search_call	OpenAI.FileSearchToolCallItemParam
computer_call	OpenAI.ComputerToolCallItemParam
computer_call_output	OpenAI.ComputerToolCallOutputItemParam
web_search_call	OpenAI.WebSearchToolCallItemParam
function_call	OpenAI.FunctionToolCallItemParam
function_call_output	OpenAI.FunctionToolCallOutputItemParam
reasoning	OpenAI.ReasoningItemParam
item_reference	OpenAI.ItemReferenceItemParam
image_generation_call	OpenAI.ImageGenToolCallItemParam
code_interpreter_call	OpenAI.CodeInterpreterToolCallItemParam
local_shell_call	OpenAI.LocalShellToolCallItemParam
local_shell_call_output	OpenAI.LocalShellToolCallOutputItemParam
mcp_list_tools	OpenAI.MCPListToolsItemParam
mcp_approval_request	OpenAI.MCPApprovalRequestItemParam
mcp_approval_response	OpenAI.MCPApprovalResponseItemParam
mcp_call	OpenAI.MCPCallItemParam
message	OpenAI.ResponsesMessageItemParam

Name	Type	Description	Required	Default
type	OpenAI.ItemType		Yes

OpenAI.ItemReferenceItemParam

An internal identifier for an item to reference.

Name	Type	Description	Required	Default
id	string	The service-originated ID of the previously generated response item being referenced.	Yes
type	enum	Possible values: item_reference	Yes

OpenAI.ItemResource

Content item used to generate a response.

Discriminator for OpenAI.ItemResource

This component uses the property type to discriminate between different types:

Type Value	Schema
file_search_call	OpenAI.FileSearchToolCallItemResource
computer_call	OpenAI.ComputerToolCallItemResource
computer_call_output	OpenAI.ComputerToolCallOutputItemResource
web_search_call	OpenAI.WebSearchToolCallItemResource
function_call	OpenAI.FunctionToolCallItemResource
function_call_output	OpenAI.FunctionToolCallOutputItemResource
reasoning	OpenAI.ReasoningItemResource
image_generation_call	OpenAI.ImageGenToolCallItemResource
code_interpreter_call	OpenAI.CodeInterpreterToolCallItemResource
local_shell_call	OpenAI.LocalShellToolCallItemResource
local_shell_call_output	OpenAI.LocalShellToolCallOutputItemResource
mcp_list_tools	OpenAI.MCPListToolsItemResource
mcp_approval_request	OpenAI.MCPApprovalRequestItemResource
mcp_approval_response	OpenAI.MCPApprovalResponseItemResource
mcp_call	OpenAI.MCPCallItemResource
message	OpenAI.ResponsesMessageItemResource

Name	Type	Description	Required	Default
id	string		Yes
type	OpenAI.ItemType		Yes

OpenAI.ItemType

Property	Value
Type	string
Values	message file_search_call function_call function_call_output computer_call computer_call_output web_search_call reasoning item_reference image_generation_call code_interpreter_call local_shell_call local_shell_call_output mcp_list_tools mcp_approval_request mcp_approval_response mcp_call

OpenAI.ListFineTuningJobCheckpointsResponse

Name	Type	Description	Required	Default
data	array		Yes
first_id	string		No
has_more	boolean		Yes
last_id	string		No
object	enum	Possible values: list	Yes

OpenAI.ListFineTuningJobEventsResponse

Name	Type	Description	Required	Default
data	array		Yes
has_more	boolean		Yes
object	enum	Possible values: list	Yes

OpenAI.ListModelsResponse

Name	Type	Description	Required	Default
data	array		Yes
object	enum	Possible values: list	Yes

OpenAI.ListPaginatedFineTuningJobsResponse

Name	Type	Description	Required	Default
data	array		Yes
has_more	boolean		Yes
object	enum	Possible values: list	Yes

OpenAI.ListVectorStoreFilesFilter

Property	Value
Type	string
Values	in_progress completed failed cancelled

OpenAI.ListVectorStoreFilesResponse

Name	Type	Description	Required	Default
data	array		Yes
first_id	string		Yes
has_more	boolean		Yes
last_id	string		Yes
object	enum	Possible values: list	Yes

OpenAI.ListVectorStoresResponse

Name	Type	Description	Required	Default
data	array		Yes
first_id	string		Yes
has_more	boolean		Yes
last_id	string		Yes
object	enum	Possible values: list	Yes

OpenAI.LocalShellExecAction

Execute a shell command on the server.

Name	Type	Description	Required	Default
command	array	The command to run.	Yes
env	object	Environment variables to set for the command.	Yes
timeout_ms	integer	Optional timeout in milliseconds for the command.	No
type	enum	The type of the local shell action. Always exec. Possible values: exec	Yes
user	string	Optional user to run the command as.	No
working_directory	string	Optional working directory to run the command in.	No

OpenAI.LocalShellTool

A tool that allows the model to execute shell commands in a local environment.

Name	Type	Description	Required	Default
type	enum	The type of the local shell tool. Always local_shell. Possible values: local_shell	Yes

OpenAI.LocalShellToolCallItemParam

A tool call to run a command on the local shell.

Name	Type	Description	Required	Default
action	OpenAI.LocalShellExecAction	Execute a shell command on the server.	Yes
call_id	string	The unique ID of the local shell tool call generated by the model.	Yes
type	enum	Possible values: local_shell_call	Yes

OpenAI.LocalShellToolCallItemResource

A tool call to run a command on the local shell.

Name	Type	Description	Required	Default
action	OpenAI.LocalShellExecAction	Execute a shell command on the server.	Yes
call_id	string	The unique ID of the local shell tool call generated by the model.	Yes
status	enum	Possible values: in_progress, completed, incomplete	Yes
type	enum	Possible values: local_shell_call	Yes

OpenAI.LocalShellToolCallOutputItemParam

The output of a local shell tool call.

Name	Type	Description	Required	Default
output	string	A JSON string of the output of the local shell tool call.	Yes
type	enum	Possible values: local_shell_call_output	Yes

OpenAI.LocalShellToolCallOutputItemResource

The output of a local shell tool call.

Name	Type	Description	Required	Default
output	string	A JSON string of the output of the local shell tool call.	Yes
status	enum	Possible values: in_progress, completed, incomplete	Yes
type	enum	Possible values: local_shell_call_output	Yes

OpenAI.Location

Discriminator for OpenAI.Location

This component uses the property type to discriminate between different types:

Type Value	Schema
approximate	OpenAI.ApproximateLocation

Name	Type	Description	Required	Default
type	OpenAI.LocationType		Yes

OpenAI.LocationType

Property	Value
Type	string
Values	approximate

OpenAI.LogProb

The log probability of a token.

Name	Type	Description	Required	Default
bytes	array		Yes
logprob	number		Yes
token	string		Yes
top_logprobs	array		Yes

OpenAI.MCPApprovalRequestItemParam

A request for human approval of a tool invocation.

Name	Type	Description	Required	Default
arguments	string	A JSON string of arguments for the tool.	Yes
name	string	The name of the tool to run.	Yes
server_label	string	The label of the MCP server making the request.	Yes
type	enum	Possible values: mcp_approval_request	Yes

OpenAI.MCPApprovalRequestItemResource

A request for human approval of a tool invocation.

Name	Type	Description	Required	Default
arguments	string	A JSON string of arguments for the tool.	Yes
name	string	The name of the tool to run.	Yes
server_label	string	The label of the MCP server making the request.	Yes
type	enum	Possible values: mcp_approval_request	Yes

OpenAI.MCPApprovalResponseItemParam

A response to an MCP approval request.

Name	Type	Description	Required	Default
approval_request_id	string	The ID of the approval request being answered.	Yes
approve	boolean	Whether the request was approved.	Yes
reason	string	Optional reason for the decision.	No
type	enum	Possible values: mcp_approval_response	Yes

OpenAI.MCPApprovalResponseItemResource

A response to an MCP approval request.

Name	Type	Description	Required	Default
approval_request_id	string	The ID of the approval request being answered.	Yes
approve	boolean	Whether the request was approved.	Yes
reason	string	Optional reason for the decision.	No
type	enum	Possible values: mcp_approval_response	Yes

OpenAI.MCPCallItemParam

An invocation of a tool on an MCP server.

Name	Type	Description	Required	Default
arguments	string	A JSON string of the arguments passed to the tool.	Yes
error	string	The error from the tool call, if any.	No
name	string	The name of the tool that was run.	Yes
output	string	The output from the tool call.	No
server_label	string	The label of the MCP server running the tool.	Yes
type	enum	Possible values: mcp_call	Yes

OpenAI.MCPCallItemResource

An invocation of a tool on an MCP server.

Name	Type	Description	Required	Default
arguments	string	A JSON string of the arguments passed to the tool.	Yes
error	string	The error from the tool call, if any.	No
name	string	The name of the tool that was run.	Yes
output	string	The output from the tool call.	No
server_label	string	The label of the MCP server running the tool.	Yes
type	enum	Possible values: mcp_call	Yes

OpenAI.MCPListToolsItemParam

A list of tools available on an MCP server.

Name	Type	Description	Required	Default
error	string	Error message if the server could not list tools.	No
server_label	string	The label of the MCP server.	Yes
tools	array	The tools available on the server.	Yes
type	enum	Possible values: mcp_list_tools	Yes

OpenAI.MCPListToolsItemResource

A list of tools available on an MCP server.

Name	Type	Description	Required	Default
error	string	Error message if the server could not list tools.	No
server_label	string	The label of the MCP server.	Yes
tools	array	The tools available on the server.	Yes
type	enum	Possible values: mcp_list_tools	Yes

OpenAI.MCPListToolsTool

A tool available on an MCP server.

Name	Type	Description	Required	Default
annotations		Additional annotations about the tool.	No
description	string	The description of the tool.	No
input_schema		The JSON schema describing the tool’s input.	Yes
name	string	The name of the tool.	Yes

OpenAI.MCPTool

Give the model access to additional tools via remote Model Context Protocol (MCP) servers.

Name	Type	Description	Required	Default
allowed_tools	object		No
└─ tool_names	array	List of allowed tool names.	No
headers	object	Optional HTTP headers to send to the MCP server. Use for authentication or other purposes.	No
require_approval	object (see valid models below)	Specify which of the MCP server’s tools require approval.	No
server_label	string	A label for this MCP server, used to identify it in tool calls.	Yes
server_url	string	The URL for the MCP server.	Yes
type	enum	The type of the MCP tool. Always mcp. Possible values: mcp	Yes

OpenAI.MetadataPropertyForRequest

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, and querying for objects via API or the dashboard.

Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

Name	Type	Description	Required	Default
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No

OpenAI.Model

Describes an OpenAI model offering that can be used with the API.

Name	Type	Description	Required	Default
created	integer	The Unix timestamp (in seconds) when the model was created.	Yes
id	string	The model identifier, which can be referenced in the API endpoints.	Yes
object	enum	The object type, which is always “model”. Possible values: model	Yes
owned_by	string	The organization that owns the model.	Yes

OpenAI.OtherChunkingStrategyResponseParam

This is returned when the chunking strategy is unknown. Typically, this is because the file was indexed before the chunking_strategy concept was introduced in the API.

Name	Type	Description	Required	Default
type	enum	Always other. Possible values: other	Yes

OpenAI.ParallelToolCalls

Whether to enable parallel function calling during tool use.

Type: boolean

OpenAI.Prompt

Reference to a prompt template and its variables.

Name	Type	Description	Required	Default
id	string	The unique identifier of the prompt template to use.	Yes
variables	object	Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.	No
version	string	Optional version of the prompt template.	No

OpenAI.RankingOptions

Name	Type	Description	Required	Default
ranker	enum	The ranker to use for the file search. Possible values: auto, default-2024-11-15	No
score_threshold	number	The score threshold for the file search, a number between 0 and 1. Numbers closer to 1 will attempt to return only the most relevant results, but may return fewer results.	No

OpenAI.Reasoning

reasoning models only

Configuration options for reasoning models.

Name	Type	Description	Required	Default
effort	object	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
generate_summary	enum	Deprecated: use summary instead. A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
summary	enum	A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No

OpenAI.ReasoningEffort

reasoning models only

Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.

Property	Value
Description	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
Type	string
Values	low medium high

OpenAI.ReasoningItemParam

A description of the chain of thought used by a reasoning model while generating a response. Be sure to include these items in your input to the Responses API for subsequent turns of a conversation if you are manually managing context.

Name	Type	Description	Required	Default
encrypted_content	string	The encrypted content of the reasoning item - populated when a response is generated with reasoning.encrypted_content in the include parameter.	No
summary	array	Reasoning text contents.	Yes
type	enum	Possible values: reasoning	Yes

OpenAI.ReasoningItemResource

A description of the chain of thought used by a reasoning model while generating a response. Be sure to include these items in your input to the Responses API for subsequent turns of a conversation if you are manually managing context.

Name	Type	Description	Required	Default
encrypted_content	string	The encrypted content of the reasoning item - populated when a response is generated with reasoning.encrypted_content in the include parameter.	No
summary	array	Reasoning text contents.	Yes
type	enum	Possible values: reasoning	Yes

OpenAI.ReasoningItemSummaryPart

Discriminator for OpenAI.ReasoningItemSummaryPart

This component uses the property type to discriminate between different types:

Type Value	Schema
summary_text	OpenAI.ReasoningItemSummaryTextPart

Name	Type	Description	Required	Default
type	OpenAI.ReasoningItemSummaryPartType		Yes

OpenAI.ReasoningItemSummaryPartType

Property	Value
Type	string
Values	summary_text

OpenAI.ReasoningItemSummaryTextPart

Name	Type	Description	Required	Default
text	string		Yes
type	enum	Possible values: summary_text	Yes

OpenAI.Response

Name	Type	Description	Required	Default
background	boolean	Whether to run the model response in the background. Learn more.	No	False
created_at	integer	Unix timestamp (in seconds) of when this Response was created.	Yes
error	object	An error object returned when the model fails to generate a Response.	Yes
└─ code	OpenAI.ResponseErrorCode	The error code for the response.	No
└─ message	string	A human-readable description of the error.	No
id	string	Unique identifier for this Response.	Yes
incomplete_details	object	Details about why the response is incomplete.	Yes
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
instructions	string or array		Yes
max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
object	enum	The object type of this resource - always set to response. Possible values: response	Yes
output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	Yes
output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	Yes	True
previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
prompt	object	Reference to a prompt template and its variables.	No
└─ id	string	The unique identifier of the prompt template to use.	No
└─ variables	OpenAI.ResponsePromptVariables	Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.	No
└─ version	string	Optional version of the prompt template.	No
reasoning	object	reasoning models only Configuration options for reasoning models.	No
└─ effort	OpenAI.ReasoningEffort	reasoning models only Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.	No
└─ generate_summary	enum	Deprecated: use summary instead. A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
└─ summary	enum	A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model’s reasoning process. One of auto, concise, or detailed. Possible values: auto, concise, detailed	No
status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
temperature	number	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. We generally recommend altering this or top_p but not both.	Yes
text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
tool_choice	object	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools.	No
└─ type	OpenAI.ToolChoiceObjectType	Indicates that the model should use a built-in tool to generate a response..	No
tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
top_logprobs	integer	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.	No
top_p	number	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. We generally recommend altering this or temperature but not both.	Yes
truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	Yes

OpenAI.ResponseCodeInterpreterCallCodeDeltaEvent

Emitted when a partial code snippet is streamed by the code interpreter.

Name	Type	Description	Required	Default
delta	string	The partial code snippet being streamed by the code interpreter.	Yes
item_id	string	The unique identifier of the code interpreter tool call item.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item in the response for which the code is being streamed.	Yes
type	enum	The type of the event. Always response.code_interpreter_call_code.delta. Possible values: response.code_interpreter_call_code.delta	Yes

OpenAI.ResponseCodeInterpreterCallCodeDoneEvent

Emitted when the code snippet is finalized by the code interpreter.

Name	Type	Description	Required	Default
code	string	The final code snippet output by the code interpreter.	Yes
item_id	string	The unique identifier of the code interpreter tool call item.	Yes
output_index	integer	The index of the output item in the response for which the code is finalized.	Yes
type	enum	The type of the event. Always response.code_interpreter_call_code.done. Possible values: response.code_interpreter_call_code.done	Yes

OpenAI.ResponseCodeInterpreterCallCompletedEvent

Emitted when the code interpreter call is completed.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the code interpreter tool call item.	Yes
output_index	integer	The index of the output item in the response for which the code interpreter call is completed.	Yes
type	enum	The type of the event. Always response.code_interpreter_call.completed. Possible values: response.code_interpreter_call.completed	Yes

OpenAI.ResponseCodeInterpreterCallInProgressEvent

Emitted when a code interpreter call is in progress.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the code interpreter tool call item.	Yes
output_index	integer	The index of the output item in the response for which the code interpreter call is in progress.	Yes
type	enum	The type of the event. Always response.code_interpreter_call.in_progress. Possible values: response.code_interpreter_call.in_progress	Yes

OpenAI.ResponseCodeInterpreterCallInterpretingEvent

Emitted when the code interpreter is actively interpreting the code snippet.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the code interpreter tool call item.	Yes
output_index	integer	The index of the output item in the response for which the code interpreter is interpreting code.	Yes
type	enum	The type of the event. Always response.code_interpreter_call.interpreting. Possible values: response.code_interpreter_call.interpreting	Yes

OpenAI.ResponseCompletedEvent

Emitted when the model response is complete.

Name	Type	Description	Required	Default
response	object		Yes
└─ background	boolean	Whether to run the model response in the background. Learn more.	No	False
└─ created_at	integer	Unix timestamp (in seconds) of when this Response was created.	No
└─ error	OpenAI.ResponseError	An error object returned when the model fails to generate a Response.	No
└─ id	string	Unique identifier for this Response.	No
└─ incomplete_details	object	Details about why the response is incomplete.	No
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
└─ instructions	string or array	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
└─ max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
└─ max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ object	enum	The object type of this resource - always set to response. Possible values: response	No
└─ output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	No
└─ output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
└─ parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
└─ previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
└─ prompt	OpenAI.Prompt	Reference to a prompt template and its variables.	No
└─ reasoning	OpenAI.Reasoning	reasoning models only Configuration options for reasoning models.	No
└─ status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
└─ temperature	number	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. We generally recommend altering this or top_p but not both.	No
└─ text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
└─ tool_choice	OpenAI.ToolChoiceOptions or OpenAI.ToolChoiceObject	How the model should select which tool (or tools) to use when generating a response. See the tools parameter to see how to specify which tools the model can call.	No
└─ tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
└─ top_logprobs	integer	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.	No
└─ top_p	number	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. We generally recommend altering this or temperature but not both.	No
└─ truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
└─ usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
└─ user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No
type	enum	The type of the event. Always response.completed. Possible values: response.completed	Yes

OpenAI.ResponseContentPartAddedEvent

Emitted when a new content part is added.

Name	Type	Description	Required	Default
content_index	integer	The index of the content part that was added.	Yes
item_id	string	The ID of the output item that the content part was added to.	Yes
output_index	integer	The index of the output item that the content part was added to.	Yes
part	object		Yes
└─ type	OpenAI.ItemContentType	Multi-modal input and output contents.	No
type	enum	The type of the event. Always response.content_part.added. Possible values: response.content_part.added	Yes

OpenAI.ResponseContentPartDoneEvent

Emitted when a content part is done.

Name	Type	Description	Required	Default
content_index	integer	The index of the content part that is done.	Yes
item_id	string	The ID of the output item that the content part was added to.	Yes
output_index	integer	The index of the output item that the content part was added to.	Yes
part	object		Yes
└─ type	OpenAI.ItemContentType	Multi-modal input and output contents.	No
type	enum	The type of the event. Always response.content_part.done. Possible values: response.content_part.done	Yes

OpenAI.ResponseCreatedEvent

An event that is emitted when a response is created.

Name	Type	Description	Required	Default
response	object		Yes
└─ background	boolean	Whether to run the model response in the background. Learn more.	No	False
└─ created_at	integer	Unix timestamp (in seconds) of when this Response was created.	No
└─ error	OpenAI.ResponseError	An error object returned when the model fails to generate a Response.	No
└─ id	string	Unique identifier for this Response.	No
└─ incomplete_details	object	Details about why the response is incomplete.	No
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
└─ instructions	string or array	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
└─ max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
└─ max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ object	enum	The object type of this resource - always set to response. Possible values: response	No
└─ output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	No
└─ output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
└─ parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
└─ previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
└─ prompt	OpenAI.Prompt	Reference to a prompt template and its variables.	No
└─ reasoning	OpenAI.Reasoning	reasoning models only Configuration options for reasoning models.	No
└─ status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
└─ temperature	number	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. We generally recommend altering this or top_p but not both.	No
└─ text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
└─ tool_choice	OpenAI.ToolChoiceOptions or OpenAI.ToolChoiceObject	How the model should select which tool (or tools) to use when generating a response. See the tools parameter to see how to specify which tools the model can call.	No
└─ tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
└─ top_logprobs	integer	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.	No
└─ top_p	number	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. We generally recommend altering this or temperature but not both.	No
└─ truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
└─ usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
└─ user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No
type	enum	The type of the event. Always response.created. Possible values: response.created	Yes

OpenAI.ResponseError

An error object returned when the model fails to generate a Response.

Name	Type	Description	Required	Default
code	OpenAI.ResponseErrorCode	The error code for the response.	Yes
message	string	A human-readable description of the error.	Yes

OpenAI.ResponseErrorCode

The error code for the response.

Property	Value
Description	The error code for the response.
Type	string
Values	server_error rate_limit_exceeded invalid_prompt vector_store_timeout invalid_image invalid_image_format invalid_base64_image invalid_image_url image_too_large image_too_small image_parse_error image_content_policy_violation invalid_image_mode image_file_too_large unsupported_image_media_type empty_image_file failed_to_download_image image_file_not_found

OpenAI.ResponseErrorEvent

Emitted when an error occurs.

Name	Type	Description	Required	Default
code	string	The error code.	Yes
message	string	The error message.	Yes
param	string	The error parameter.	Yes
type	enum	The type of the event. Always error. Possible values: error	Yes

OpenAI.ResponseFailedEvent

An event that is emitted when a response fails.

Name	Type	Description	Required	Default
response	object		Yes
└─ background	boolean	Whether to run the model response in the background. Learn more.	No	False
└─ created_at	integer	Unix timestamp (in seconds) of when this Response was created.	No
└─ error	OpenAI.ResponseError	An error object returned when the model fails to generate a Response.	No
└─ id	string	Unique identifier for this Response.	No
└─ incomplete_details	object	Details about why the response is incomplete.	No
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
└─ instructions	string or array	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
└─ max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
└─ max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ object	enum	The object type of this resource - always set to response. Possible values: response	No
└─ output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	No
└─ output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
└─ parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
└─ previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
└─ prompt	OpenAI.Prompt	Reference to a prompt template and its variables.	No
└─ reasoning	OpenAI.Reasoning	reasoning models only Configuration options for reasoning models.	No
└─ status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
└─ temperature	number	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. We generally recommend altering this or top_p but not both.	No
└─ text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
└─ tool_choice	OpenAI.ToolChoiceOptions or OpenAI.ToolChoiceObject	How the model should select which tool (or tools) to use when generating a response. See the tools parameter to see how to specify which tools the model can call.	No
└─ tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
└─ top_logprobs	integer	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.	No
└─ top_p	number	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. We generally recommend altering this or temperature but not both.	No
└─ truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
└─ usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
└─ user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No
type	enum	The type of the event. Always response.failed. Possible values: response.failed	Yes

OpenAI.ResponseFileSearchCallCompletedEvent

Emitted when a file search call is completed (results found).

Name	Type	Description	Required	Default
item_id	string	The ID of the output item that the file search call is initiated.	Yes
output_index	integer	The index of the output item that the file search call is initiated.	Yes
type	enum	The type of the event. Always response.file_search_call.completed. Possible values: response.file_search_call.completed	Yes

OpenAI.ResponseFileSearchCallInProgressEvent

Emitted when a file search call is initiated.

Name	Type	Description	Required	Default
item_id	string	The ID of the output item that the file search call is initiated.	Yes
output_index	integer	The index of the output item that the file search call is initiated.	Yes
type	enum	The type of the event. Always response.file_search_call.in_progress. Possible values: response.file_search_call.in_progress	Yes

OpenAI.ResponseFileSearchCallSearchingEvent

Emitted when a file search is currently searching.

Name	Type	Description	Required	Default
item_id	string	The ID of the output item that the file search call is initiated.	Yes
output_index	integer	The index of the output item that the file search call is searching.	Yes
type	enum	The type of the event. Always response.file_search_call.searching. Possible values: response.file_search_call.searching	Yes

OpenAI.ResponseFormat

Discriminator for OpenAI.ResponseFormat

This component uses the property type to discriminate between different types:

Type Value	Schema
text	OpenAI.ResponseFormatText
json_object	OpenAI.ResponseFormatJsonObject
json_schema	OpenAI.ResponseFormatJsonSchema

Name	Type	Description	Required	Default
type	enum	Possible values: text, json_object, json_schema	Yes

OpenAI.ResponseFormatJsonObject

JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

Name	Type	Description	Required	Default
type	enum	The type of response format being defined. Always json_object. Possible values: json_object	Yes

OpenAI.ResponseFormatJsonSchema

JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

Name	Type	Description	Required	Default
json_schema	object	Structured Outputs configuration options, including a JSON Schema.	Yes
└─ description	string	A description of what the response format is for, used by the model to determine how to respond in the format.	No
└─ name	string	The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.	No
└─ schema	OpenAI.ResponseFormatJsonSchemaSchema	The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.	No
└─ strict	boolean	Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.	No	False
type	enum	The type of response format being defined. Always json_schema. Possible values: json_schema	Yes

OpenAI.ResponseFormatJsonSchemaSchema

The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

Type: object

OpenAI.ResponseFormatText

Default response format. Used to generate text responses.

Name	Type	Description	Required	Default
type	enum	The type of response format being defined. Always text. Possible values: text	Yes

OpenAI.ResponseFunctionCallArgumentsDeltaEvent

Emitted when there is a partial function-call arguments delta.

Name	Type	Description	Required	Default
delta	string	The function-call arguments delta that is added.	Yes
item_id	string	The ID of the output item that the function-call arguments delta is added to.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item that the function-call arguments delta is added to.	Yes
type	enum	The type of the event. Always response.function_call_arguments.delta. Possible values: response.function_call_arguments.delta	Yes

OpenAI.ResponseFunctionCallArgumentsDoneEvent

Emitted when function-call arguments are finalized.

Name	Type	Description	Required	Default
arguments	string	The function-call arguments.	Yes
item_id	string	The ID of the item.	Yes
output_index	integer	The index of the output item.	Yes
type	enum	Possible values: response.function_call_arguments.done	Yes

OpenAI.ResponseImageGenCallCompletedEvent

Emitted when an image generation tool call has completed and the final image is available.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the image generation item being processed.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.image_generation_call.completed’. Possible values: response.image_generation_call.completed	Yes

OpenAI.ResponseImageGenCallGeneratingEvent

Emitted when an image generation tool call is actively generating an image (intermediate state).

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the image generation item being processed.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.image_generation_call.generating’. Possible values: response.image_generation_call.generating	Yes

OpenAI.ResponseImageGenCallInProgressEvent

Emitted when an image generation tool call is in progress.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the image generation item being processed.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.image_generation_call.in_progress’. Possible values: response.image_generation_call.in_progress	Yes

OpenAI.ResponseImageGenCallPartialImageEvent

Emitted when a partial image is available during image generation streaming.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the image generation item being processed.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
partial_image_b64	string	Base64-encoded partial image data, suitable for rendering as an image.	Yes
partial_image_index	integer	0-based index for the partial image (backend is 1-based, but this is 0-based for the user).	Yes
type	enum	The type of the event. Always ‘response.image_generation_call.partial_image’. Possible values: response.image_generation_call.partial_image	Yes

OpenAI.ResponseInProgressEvent

Emitted when the response is in progress.

Name	Type	Description	Required	Default
response	object		Yes
└─ background	boolean	Whether to run the model response in the background. Learn more.	No	False
└─ created_at	integer	Unix timestamp (in seconds) of when this Response was created.	No
└─ error	OpenAI.ResponseError	An error object returned when the model fails to generate a Response.	No
└─ id	string	Unique identifier for this Response.	No
└─ incomplete_details	object	Details about why the response is incomplete.	No
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
└─ instructions	string or array	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
└─ max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
└─ max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ object	enum	The object type of this resource - always set to response. Possible values: response	No
└─ output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	No
└─ output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
└─ parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
└─ previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
└─ prompt	OpenAI.Prompt	Reference to a prompt template and its variables.	No
└─ reasoning	OpenAI.Reasoning	reasoning models only Configuration options for reasoning models.	No
└─ status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
└─ temperature	number	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. We generally recommend altering this or top_p but not both.	No
└─ text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
└─ tool_choice	OpenAI.ToolChoiceOptions or OpenAI.ToolChoiceObject	How the model should select which tool (or tools) to use when generating a response. See the tools parameter to see how to specify which tools the model can call.	No
└─ tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
└─ top_logprobs	integer	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.	No
└─ top_p	number	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. We generally recommend altering this or temperature but not both.	No
└─ truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
└─ usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
└─ user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No
type	enum	The type of the event. Always response.in_progress. Possible values: response.in_progress	Yes

OpenAI.ResponseIncompleteEvent

An event that is emitted when a response finishes as incomplete.

Name	Type	Description	Required	Default
response	object		Yes
└─ background	boolean	Whether to run the model response in the background. Learn more.	No	False
└─ created_at	integer	Unix timestamp (in seconds) of when this Response was created.	No
└─ error	OpenAI.ResponseError	An error object returned when the model fails to generate a Response.	No
└─ id	string	Unique identifier for this Response.	No
└─ incomplete_details	object	Details about why the response is incomplete.	No
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
└─ instructions	string or array	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
└─ max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
└─ max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ object	enum	The object type of this resource - always set to response. Possible values: response	No
└─ output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	No
└─ output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
└─ parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
└─ previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
└─ prompt	OpenAI.Prompt	Reference to a prompt template and its variables.	No
└─ reasoning	OpenAI.Reasoning	reasoning models only Configuration options for reasoning models.	No
└─ status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
└─ temperature	number	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. We generally recommend altering this or top_p but not both.	No
└─ text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
└─ tool_choice	OpenAI.ToolChoiceOptions or OpenAI.ToolChoiceObject	How the model should select which tool (or tools) to use when generating a response. See the tools parameter to see how to specify which tools the model can call.	No
└─ tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
└─ top_logprobs	integer	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.	No
└─ top_p	number	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. We generally recommend altering this or temperature but not both.	No
└─ truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
└─ usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
└─ user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No
type	enum	The type of the event. Always response.incomplete. Possible values: response.incomplete	Yes

OpenAI.ResponseItemList

A list of Response items.

Name	Type	Description	Required	Default
data	array	A list of items used to generate this response.	Yes
first_id	string	The ID of the first item in the list.	Yes
has_more	boolean	Whether there are more items available.	Yes
last_id	string	The ID of the last item in the list.	Yes
object	enum	The type of object returned, must be list. Possible values: list	Yes

OpenAI.ResponseMCPCallArgumentsDeltaEvent

Emitted when there is a delta (partial update) to the arguments of an MCP tool call.

Name	Type	Description	Required	Default
delta		The partial update to the arguments for the MCP tool call.	Yes
item_id	string	The unique identifier of the MCP tool call item being processed.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.mcp_call.arguments_delta’. Possible values: response.mcp_call.arguments_delta	Yes

OpenAI.ResponseMCPCallArgumentsDoneEvent

Emitted when the arguments for an MCP tool call are finalized.

Name	Type	Description	Required	Default
arguments		The finalized arguments for the MCP tool call.	Yes
item_id	string	The unique identifier of the MCP tool call item being processed.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.mcp_call.arguments_done’. Possible values: response.mcp_call.arguments_done	Yes

OpenAI.ResponseMCPCallCompletedEvent

Emitted when an MCP tool call has completed successfully.

Name	Type	Description	Required	Default
type	enum	The type of the event. Always ‘response.mcp_call.completed’. Possible values: response.mcp_call.completed	Yes

OpenAI.ResponseMCPCallFailedEvent

Emitted when an MCP tool call has failed.

Name	Type	Description	Required	Default
type	enum	The type of the event. Always ‘response.mcp_call.failed’. Possible values: response.mcp_call.failed	Yes

OpenAI.ResponseMCPCallInProgressEvent

Emitted when an MCP tool call is in progress.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the MCP tool call item being processed.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.mcp_call.in_progress’. Possible values: response.mcp_call.in_progress	Yes

OpenAI.ResponseMCPListToolsCompletedEvent

Emitted when the list of available MCP tools has been successfully retrieved.

Name	Type	Description	Required	Default
type	enum	The type of the event. Always ‘response.mcp_list_tools.completed’. Possible values: response.mcp_list_tools.completed	Yes

OpenAI.ResponseMCPListToolsFailedEvent

Emitted when the attempt to list available MCP tools has failed.

Name	Type	Description	Required	Default
type	enum	The type of the event. Always ‘response.mcp_list_tools.failed’. Possible values: response.mcp_list_tools.failed	Yes

OpenAI.ResponseMCPListToolsInProgressEvent

Emitted when the system is in the process of retrieving the list of available MCP tools.

Name	Type	Description	Required	Default
type	enum	The type of the event. Always ‘response.mcp_list_tools.in_progress’. Possible values: response.mcp_list_tools.in_progress	Yes

OpenAI.ResponseOutputItemAddedEvent

Emitted when a new output item is added.

Name	Type	Description	Required	Default
item	object	Content item used to generate a response.	Yes
└─ id	string		No
└─ type	OpenAI.ItemType		No
output_index	integer	The index of the output item that was added.	Yes
type	enum	The type of the event. Always response.output_item.added. Possible values: response.output_item.added	Yes

OpenAI.ResponseOutputItemDoneEvent

Emitted when an output item is marked done.

Name	Type	Description	Required	Default
item	object	Content item used to generate a response.	Yes
└─ id	string		No
└─ type	OpenAI.ItemType		No
output_index	integer	The index of the output item that was marked done.	Yes
type	enum	The type of the event. Always response.output_item.done. Possible values: response.output_item.done	Yes

OpenAI.ResponsePromptVariables

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

Type: object

OpenAI.ResponseQueuedEvent

Emitted when a response is queued and waiting to be processed.

Name	Type	Description	Required	Default
response	object		Yes
└─ background	boolean	Whether to run the model response in the background. Learn more.	No	False
└─ created_at	integer	Unix timestamp (in seconds) of when this Response was created.	No
└─ error	OpenAI.ResponseError	An error object returned when the model fails to generate a Response.	No
└─ id	string	Unique identifier for this Response.	No
└─ incomplete_details	object	Details about why the response is incomplete.	No
└─ reason	enum	The reason why the response is incomplete. Possible values: max_output_tokens, content_filter	No
└─ instructions	string or array	A system (or developer) message inserted into the model’s context. When using along with previous_response_id, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses.	No
└─ max_output_tokens	integer	An upper bound for the number of tokens that can be generated for a response, including visible output tokens and reasoning tokens.	No
└─ max_tool_calls	integer	The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored.	No
└─ metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
└─ object	enum	The object type of this resource - always set to response. Possible values: response	No
└─ output	array	An array of content items generated by the model. - The length and order of items in the output array is dependent on the model’s response. - Rather than accessing the first item in the output array and assuming it’s an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.	No
└─ output_text	string	SDK-only convenience property that contains the aggregated text output from all output_text items in the output array, if any are present. Supported in the Python and JavaScript SDKs.	No
└─ parallel_tool_calls	boolean	Whether to allow the model to run tool calls in parallel.	No	True
└─ previous_response_id	string	The unique ID of the previous response to the model. Use this to create multi-turn conversations.	No
└─ prompt	OpenAI.Prompt	Reference to a prompt template and its variables.	No
└─ reasoning	OpenAI.Reasoning	reasoning models only Configuration options for reasoning models.	No
└─ status	enum	The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete. Possible values: completed, failed, in_progress, cancelled, queued, incomplete	No
└─ temperature	number	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. We generally recommend altering this or top_p but not both.	No
└─ text	object	Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: Structured Outputs	No
└─ format	OpenAI.ResponseTextFormatConfiguration		No
└─ tool_choice	OpenAI.ToolChoiceOptions or OpenAI.ToolChoiceObject	How the model should select which tool (or tools) to use when generating a response. See the tools parameter to see how to specify which tools the model can call.	No
└─ tools	array	An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. The two categories of tools you can provide the model are: - Built-in tools: Tools that are provided by OpenAI that extend the model’s capabilities, like web search or file search.	No
└─ top_logprobs	integer	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.	No
└─ top_p	number	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. We generally recommend altering this or temperature but not both.	No
└─ truncation	enum	The truncation strategy to use for the model response. - auto: If the context of this response and previous ones exceeds the model’s context window size, the model will truncate the response to fit the context window by dropping input items in the middle of the conversation. - disabled (default): If a model response will exceed the context window size for a model, the request will fail with a 400 error. Possible values: auto, disabled	No
└─ usage	OpenAI.ResponseUsage	Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.	No
└─ user	string	A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.	No
type	enum	The type of the event. Always ‘response.queued’. Possible values: response.queued	Yes

OpenAI.ResponseReasoningDeltaEvent

Emitted when there is a delta (partial update) to the reasoning content.

Name	Type	Description	Required	Default
content_index	integer	The index of the reasoning content part within the output item.	Yes
delta		The partial update to the reasoning content.	Yes
item_id	string	The unique identifier of the item for which reasoning is being updated.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
type	enum	The type of the event. Always ‘response.reasoning.delta’. Possible values: response.reasoning.delta	Yes

OpenAI.ResponseReasoningDoneEvent

Emitted when the reasoning content is finalized for an item.

Name	Type	Description	Required	Default
content_index	integer	The index of the reasoning content part within the output item.	Yes
item_id	string	The unique identifier of the item for which reasoning is finalized.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
text	string	The finalized reasoning text.	Yes
type	enum	The type of the event. Always ‘response.reasoning.done’. Possible values: response.reasoning.done	Yes

OpenAI.ResponseReasoningSummaryDeltaEvent

Emitted when there is a delta (partial update) to the reasoning summary content.

Name	Type	Description	Required	Default
delta		The partial update to the reasoning summary content.	Yes
item_id	string	The unique identifier of the item for which the reasoning summary is being updated.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
summary_index	integer	The index of the summary part within the output item.	Yes
type	enum	The type of the event. Always ‘response.reasoning_summary.delta’. Possible values: response.reasoning_summary.delta	Yes

OpenAI.ResponseReasoningSummaryDoneEvent

Emitted when the reasoning summary content is finalized for an item.

Name	Type	Description	Required	Default
item_id	string	The unique identifier of the item for which the reasoning summary is finalized.	Yes
output_index	integer	The index of the output item in the response’s output array.	Yes
summary_index	integer	The index of the summary part within the output item.	Yes
text	string	The finalized reasoning summary text.	Yes
type	enum	The type of the event. Always ‘response.reasoning_summary.done’. Possible values: response.reasoning_summary.done	Yes

OpenAI.ResponseReasoningSummaryPartAddedEvent

Emitted when a new reasoning summary part is added.

Name	Type	Description	Required	Default
item_id	string	The ID of the item this summary part is associated with.	Yes
output_index	integer	The index of the output item this summary part is associated with.	Yes
part	object		Yes
└─ type	OpenAI.ReasoningItemSummaryPartType		No
summary_index	integer	The index of the summary part within the reasoning summary.	Yes
type	enum	The type of the event. Always response.reasoning_summary_part.added. Possible values: response.reasoning_summary_part.added	Yes

OpenAI.ResponseReasoningSummaryPartDoneEvent

Emitted when a reasoning summary part is completed.

Name	Type	Description	Required	Default
item_id	string	The ID of the item this summary part is associated with.	Yes
output_index	integer	The index of the output item this summary part is associated with.	Yes
part	object		Yes
└─ type	OpenAI.ReasoningItemSummaryPartType		No
summary_index	integer	The index of the summary part within the reasoning summary.	Yes
type	enum	The type of the event. Always response.reasoning_summary_part.done. Possible values: response.reasoning_summary_part.done	Yes

OpenAI.ResponseReasoningSummaryTextDeltaEvent

Emitted when a delta is added to a reasoning summary text.

Name	Type	Description	Required	Default
delta	string	The text delta that was added to the summary.	Yes
item_id	string	The ID of the item this summary text delta is associated with.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item this summary text delta is associated with.	Yes
summary_index	integer	The index of the summary part within the reasoning summary.	Yes
type	enum	The type of the event. Always response.reasoning_summary_text.delta. Possible values: response.reasoning_summary_text.delta	Yes

OpenAI.ResponseReasoningSummaryTextDoneEvent

Emitted when a reasoning summary text is completed.

Name	Type	Description	Required	Default
item_id	string	The ID of the item this summary text is associated with.	Yes
output_index	integer	The index of the output item this summary text is associated with.	Yes
summary_index	integer	The index of the summary part within the reasoning summary.	Yes
text	string	The full text of the completed reasoning summary.	Yes
type	enum	The type of the event. Always response.reasoning_summary_text.done. Possible values: response.reasoning_summary_text.done	Yes

OpenAI.ResponseRefusalDeltaEvent

Emitted when there is a partial refusal text.

Name	Type	Description	Required	Default
content_index	integer	The index of the content part that the refusal text is added to.	Yes
delta	string	The refusal text that is added.	Yes
item_id	string	The ID of the output item that the refusal text is added to.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item that the refusal text is added to.	Yes
type	enum	The type of the event. Always response.refusal.delta. Possible values: response.refusal.delta	Yes

OpenAI.ResponseRefusalDoneEvent

Emitted when refusal text is finalized.

Name	Type	Description	Required	Default
content_index	integer	The index of the content part that the refusal text is finalized.	Yes
item_id	string	The ID of the output item that the refusal text is finalized.	Yes
output_index	integer	The index of the output item that the refusal text is finalized.	Yes
refusal	string	The refusal text that is finalized.	Yes
type	enum	The type of the event. Always response.refusal.done. Possible values: response.refusal.done	Yes

OpenAI.ResponseStreamEvent

Discriminator for OpenAI.ResponseStreamEvent

This component uses the property type to discriminate between different types:

Type Value	Schema
response.completed	OpenAI.ResponseCompletedEvent
response.content_part.added	OpenAI.ResponseContentPartAddedEvent
response.content_part.done	OpenAI.ResponseContentPartDoneEvent
response.created	OpenAI.ResponseCreatedEvent
error	OpenAI.ResponseErrorEvent
response.file_search_call.completed	OpenAI.ResponseFileSearchCallCompletedEvent
response.file_search_call.in_progress	OpenAI.ResponseFileSearchCallInProgressEvent
response.file_search_call.searching	OpenAI.ResponseFileSearchCallSearchingEvent
response.function_call_arguments.delta	OpenAI.ResponseFunctionCallArgumentsDeltaEvent
response.function_call_arguments.done	OpenAI.ResponseFunctionCallArgumentsDoneEvent
response.in_progress	OpenAI.ResponseInProgressEvent
response.failed	OpenAI.ResponseFailedEvent
response.incomplete	OpenAI.ResponseIncompleteEvent
response.output_item.added	OpenAI.ResponseOutputItemAddedEvent
response.output_item.done	OpenAI.ResponseOutputItemDoneEvent
response.refusal.delta	OpenAI.ResponseRefusalDeltaEvent
response.refusal.done	OpenAI.ResponseRefusalDoneEvent
response.output_text.delta	OpenAI.ResponseTextDeltaEvent
response.output_text.done	OpenAI.ResponseTextDoneEvent
response.reasoning_summary_part.added	OpenAI.ResponseReasoningSummaryPartAddedEvent
response.reasoning_summary_part.done	OpenAI.ResponseReasoningSummaryPartDoneEvent
response.reasoning_summary_text.delta	OpenAI.ResponseReasoningSummaryTextDeltaEvent
response.reasoning_summary_text.done	OpenAI.ResponseReasoningSummaryTextDoneEvent
response.web_search_call.completed	OpenAI.ResponseWebSearchCallCompletedEvent
response.web_search_call.in_progress	OpenAI.ResponseWebSearchCallInProgressEvent
response.web_search_call.searching	OpenAI.ResponseWebSearchCallSearchingEvent
response.image_generation_call.completed	OpenAI.ResponseImageGenCallCompletedEvent
response.image_generation_call.generating	OpenAI.ResponseImageGenCallGeneratingEvent
response.image_generation_call.in_progress	OpenAI.ResponseImageGenCallInProgressEvent
response.image_generation_call.partial_image	OpenAI.ResponseImageGenCallPartialImageEvent
response.mcp_call.arguments_delta	OpenAI.ResponseMCPCallArgumentsDeltaEvent
response.mcp_call.arguments_done	OpenAI.ResponseMCPCallArgumentsDoneEvent
response.mcp_call.completed	OpenAI.ResponseMCPCallCompletedEvent
response.mcp_call.failed	OpenAI.ResponseMCPCallFailedEvent
response.mcp_call.in_progress	OpenAI.ResponseMCPCallInProgressEvent
response.mcp_list_tools.completed	OpenAI.ResponseMCPListToolsCompletedEvent
response.mcp_list_tools.failed	OpenAI.ResponseMCPListToolsFailedEvent
response.mcp_list_tools.in_progress	OpenAI.ResponseMCPListToolsInProgressEvent
response.queued	OpenAI.ResponseQueuedEvent
response.reasoning.delta	OpenAI.ResponseReasoningDeltaEvent
response.reasoning.done	OpenAI.ResponseReasoningDoneEvent
response.reasoning_summary.delta	OpenAI.ResponseReasoningSummaryDeltaEvent
response.reasoning_summary.done	OpenAI.ResponseReasoningSummaryDoneEvent
response.code_interpreter_call_code.delta	OpenAI.ResponseCodeInterpreterCallCodeDeltaEvent
response.code_interpreter_call_code.done	OpenAI.ResponseCodeInterpreterCallCodeDoneEvent
response.code_interpreter_call.completed	OpenAI.ResponseCodeInterpreterCallCompletedEvent
response.code_interpreter_call.in_progress	OpenAI.ResponseCodeInterpreterCallInProgressEvent
response.code_interpreter_call.interpreting	OpenAI.ResponseCodeInterpreterCallInterpretingEvent

Name	Type	Description	Required	Default
sequence_number	integer	The sequence number for this event.	Yes
type	OpenAI.ResponseStreamEventType		Yes

OpenAI.ResponseStreamEventType

Property	Value
Type	string
Values	response.audio.delta response.audio.done response.audio_transcript.delta response.audio_transcript.done response.code_interpreter_call_code.delta response.code_interpreter_call_code.done response.code_interpreter_call.completed response.code_interpreter_call.in_progress response.code_interpreter_call.interpreting response.completed response.content_part.added response.content_part.done response.created error response.file_search_call.completed response.file_search_call.in_progress response.file_search_call.searching response.function_call_arguments.delta response.function_call_arguments.done response.in_progress response.failed response.incomplete response.output_item.added response.output_item.done response.refusal.delta response.refusal.done response.output_text.annotation.added response.output_text.delta response.output_text.done response.reasoning_summary_part.added response.reasoning_summary_part.done response.reasoning_summary_text.delta response.reasoning_summary_text.done response.web_search_call.completed response.web_search_call.in_progress response.web_search_call.searching response.image_generation_call.completed response.image_generation_call.generating response.image_generation_call.in_progress response.image_generation_call.partial_image response.mcp_call.arguments_delta response.mcp_call.arguments_done response.mcp_call.completed response.mcp_call.failed response.mcp_call.in_progress response.mcp_list_tools.completed response.mcp_list_tools.failed response.mcp_list_tools.in_progress response.queued response.reasoning.delta response.reasoning.done response.reasoning_summary.delta response.reasoning_summary.done

OpenAI.ResponseTextDeltaEvent

Emitted when there is an additional text delta.

Name	Type	Description	Required	Default
content_index	integer	The index of the content part that the text delta was added to.	Yes
delta	string	The text delta that was added.	Yes
item_id	string	The ID of the output item that the text delta was added to.	Yes
obfuscation	string	A field of random characters introduced by stream obfuscation. Stream obfuscation is a mechanism that mitigates certain side-channel attacks.	Yes
output_index	integer	The index of the output item that the text delta was added to.	Yes
type	enum	The type of the event. Always response.output_text.delta. Possible values: response.output_text.delta	Yes

OpenAI.ResponseTextDoneEvent

Emitted when text content is finalized.

Name	Type	Description	Required	Default
content_index	integer	The index of the content part that the text content is finalized.	Yes
item_id	string	The ID of the output item that the text content is finalized.	Yes
output_index	integer	The index of the output item that the text content is finalized.	Yes
text	string	The text content that is finalized.	Yes
type	enum	The type of the event. Always response.output_text.done. Possible values: response.output_text.done	Yes

OpenAI.ResponseTextFormatConfiguration

Discriminator for OpenAI.ResponseTextFormatConfiguration

This component uses the property type to discriminate between different types:

Type Value	Schema
text	OpenAI.ResponseTextFormatConfigurationText
json_object	OpenAI.ResponseTextFormatConfigurationJsonObject
json_schema	OpenAI.ResponseTextFormatConfigurationJsonSchema

Name	Type	Description	Required	Default
type	OpenAI.ResponseTextFormatConfigurationType	An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models: Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.	Yes

OpenAI.ResponseTextFormatConfigurationJsonObject

Name	Type	Description	Required	Default
type	enum	Possible values: json_object	Yes

OpenAI.ResponseTextFormatConfigurationJsonSchema

JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

Name	Type	Description	Required	Default
description	string	A description of what the response format is for, used by the model to determine how to respond in the format.	No
name	string	The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.	Yes
schema	OpenAI.ResponseFormatJsonSchemaSchema	The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.	Yes
strict	boolean	Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.	No	False
type	enum	The type of response format being defined. Always json_schema. Possible values: json_schema	Yes

OpenAI.ResponseTextFormatConfigurationText

Name	Type	Description	Required	Default
type	enum	Possible values: text	Yes

OpenAI.ResponseTextFormatConfigurationType

An object specifying the format that the model must output.

Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

The default format is { "type": "text" } with no additional options.

Not recommended for gpt-4o and newer models:

Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

Property	Value
Description	An object specifying the format that the model must output.

Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

The default format is { "type": "text" } with no additional options.

Not recommended for gpt-4o and newer models:

Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it. | | Type | string | | Values | text
json_schema
json_object |

OpenAI.ResponseUsage

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

Name	Type	Description	Required	Default
input_tokens	integer	The number of input tokens.	Yes
input_tokens_details	object	A detailed breakdown of the input tokens.	Yes
└─ cached_tokens	integer	The number of tokens that were retrieved from the cache. More on prompt caching.	No
output_tokens	integer	The number of output tokens.	Yes
output_tokens_details	object	A detailed breakdown of the output tokens.	Yes
└─ reasoning_tokens	integer	The number of reasoning tokens.	No
total_tokens	integer	The total number of tokens used.	Yes

OpenAI.ResponseWebSearchCallCompletedEvent

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
item_id	string	Unique ID for the output item associated with the web search call.	Yes
output_index	integer	The index of the output item that the web search call is associated with.	Yes
type	enum	The type of the event. Always response.web_search_call.completed. Possible values: response.web_search_call.completed	Yes

OpenAI.ResponseWebSearchCallInProgressEvent

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
item_id	string	Unique ID for the output item associated with the web search call.	Yes
output_index	integer	The index of the output item that the web search call is associated with.	Yes
type	enum	The type of the event. Always response.web_search_call.in_progress. Possible values: response.web_search_call.in_progress	Yes

OpenAI.ResponseWebSearchCallSearchingEvent

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
item_id	string	Unique ID for the output item associated with the web search call.	Yes
output_index	integer	The index of the output item that the web search call is associated with.	Yes
type	enum	The type of the event. Always response.web_search_call.searching. Possible values: response.web_search_call.searching	Yes

OpenAI.ResponsesAssistantMessageItemParam

A message parameter item with the assistant role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always assistant. Possible values: assistant	Yes

OpenAI.ResponsesAssistantMessageItemResource

A message resource item with the assistant role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always assistant. Possible values: assistant	Yes

OpenAI.ResponsesDeveloperMessageItemParam

A message parameter item with the developer role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always developer. Possible values: developer	Yes

OpenAI.ResponsesDeveloperMessageItemResource

A message resource item with the developer role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always developer. Possible values: developer	Yes

OpenAI.ResponsesMessageItemParam

A response message item, representing a role and content, as provided as client request parameters.

Discriminator for OpenAI.ResponsesMessageItemParam

This component uses the property role to discriminate between different types:

Type Value	Schema
user	OpenAI.ResponsesUserMessageItemParam
system	OpenAI.ResponsesSystemMessageItemParam
developer	OpenAI.ResponsesDeveloperMessageItemParam
assistant	OpenAI.ResponsesAssistantMessageItemParam

Name	Type	Description	Required	Default
role	object	The collection of valid roles for responses message items.	Yes
type	enum	The type of the responses item, which is always ‘message’. Possible values: message	Yes

OpenAI.ResponsesMessageItemResource

A response message resource item, representing a role and content, as provided on service responses.

Discriminator for OpenAI.ResponsesMessageItemResource

This component uses the property role to discriminate between different types:

Type Value	Schema
user	OpenAI.ResponsesUserMessageItemResource
system	OpenAI.ResponsesSystemMessageItemResource
developer	OpenAI.ResponsesDeveloperMessageItemResource
assistant	OpenAI.ResponsesAssistantMessageItemResource

Name	Type	Description	Required	Default
role	object	The collection of valid roles for responses message items.	Yes
status	enum	The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API. Possible values: in_progress, completed, incomplete	Yes
type	enum	The type of the responses item, which is always ‘message’. Possible values: message	Yes

OpenAI.ResponsesMessageRole

The collection of valid roles for responses message items.

Property	Value
Description	The collection of valid roles for responses message items.
Type	string
Values	system developer user assistant

OpenAI.ResponsesSystemMessageItemParam

A message parameter item with the system role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always system. Possible values: system	Yes

OpenAI.ResponsesSystemMessageItemResource

A message resource item with the system role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always system. Possible values: system	Yes

OpenAI.ResponsesUserMessageItemParam

A message parameter item with the user role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always user. Possible values: user	Yes

OpenAI.ResponsesUserMessageItemResource

A message resource item with the user role.

Name	Type	Description	Required	Default
content	array	The content associated with the message.	Yes
role	enum	The role of the message, which is always user. Possible values: user	Yes

OpenAI.RunGraderRequest

Name	Type	Description	Required	Default
grader	object	A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.	Yes
└─ calculate_output	string	A formula to calculate the output based on grader results.	No
└─ evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	No
└─ graders	object		No
└─ image_tag	string	The image tag to use for the python script.	No
└─ input	array	The input text. This may include template strings.	No
└─ model	string	The model to use for the evaluation.	No
└─ name	string	The name of the grader.	No
└─ operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	No
└─ range	array	The range of the score. Defaults to [0, 1].	No
└─ reference	string	The text being graded against.	No
└─ sampling_params		The sampling parameters for the model.	No
└─ source	string	The source code of the python script.	No
└─ type	enum	The object type, which is always multi. Possible values: multi	No
item		The dataset item provided to the grader. This will be used to populate the item namespace. See the guide for more details.	No
model_sample	string	The model sample to be evaluated. This value will be used to populate the sample namespace. See the guide for more details. The output_json variable will be populated if the model sample is a valid JSON string.	Yes

OpenAI.RunGraderResponse

Name	Type	Description	Required	Default
metadata	object		Yes
└─ errors	object		No
└─ formula_parse_error	boolean		No
└─ invalid_variable_error	boolean		No
└─ model_grader_parse_error	boolean		No
└─ model_grader_refusal_error	boolean		No
└─ model_grader_server_error	boolean		No
└─ model_grader_server_error_details	string		No
└─ other_error	boolean		No
└─ python_grader_runtime_error	boolean		No
└─ python_grader_runtime_error_details	string		No
└─ python_grader_server_error	boolean		No
└─ python_grader_server_error_type	string		No
└─ sample_parse_error	boolean		No
└─ truncated_observation_error	boolean		No
└─ unresponsive_reward_error	boolean		No
└─ execution_time	number		No
└─ name	string		No
└─ sampled_model_name	string		No
└─ scores			No
└─ token_usage	integer		No
└─ type	string		No
model_grader_token_usage_per_model			Yes
reward	number		Yes
sub_rewards			Yes

OpenAI.StaticChunkingStrategy

Name	Type	Description	Required	Default
chunk_overlap_tokens	integer	The number of tokens that overlap between chunks. The default value is 400. Note that the overlap must not exceed half of max_chunk_size_tokens.	Yes
max_chunk_size_tokens	integer	The maximum number of tokens in each chunk. The default value is 800. The minimum value is 100 and the maximum value is 4096.	Yes

OpenAI.StaticChunkingStrategyRequestParam

Customize your own chunking strategy by setting chunk size and chunk overlap.

Name	Type	Description	Required	Default
static	OpenAI.StaticChunkingStrategy		Yes
type	enum	Always static. Possible values: static	Yes

OpenAI.StaticChunkingStrategyResponseParam

Name	Type	Description	Required	Default
static	OpenAI.StaticChunkingStrategy		Yes
type	enum	Always static. Possible values: static	Yes

OpenAI.StopConfiguration

Not supported with latest reasoning models o3 and o4-mini.

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

This schema accepts one of the following types:

• string
• array

OpenAI.Tool

Discriminator for OpenAI.Tool

This component uses the property type to discriminate between different types:

Type Value	Schema
function	OpenAI.FunctionTool
file_search	OpenAI.FileSearchTool
computer_use_preview	OpenAI.ComputerUsePreviewTool
web_search_preview	OpenAI.WebSearchPreviewTool
code_interpreter	OpenAI.CodeInterpreterTool
image_generation	OpenAI.ImageGenTool
local_shell	OpenAI.LocalShellTool
mcp	OpenAI.MCPTool

Name	Type	Description	Required	Default
type	OpenAI.ToolType	A tool that can be used to generate a response.	Yes

OpenAI.ToolChoiceObject

Discriminator for OpenAI.ToolChoiceObject

This component uses the property type to discriminate between different types:

Type Value	Schema
file_search	OpenAI.ToolChoiceObjectFileSearch
computer_use_preview	OpenAI.ToolChoiceObjectComputer
web_search_preview	OpenAI.ToolChoiceObjectWebSearch
image_generation	OpenAI.ToolChoiceObjectImageGen
code_interpreter	OpenAI.ToolChoiceObjectCodeInterpreter
function	OpenAI.ToolChoiceObjectFunction
mcp	OpenAI.ToolChoiceObjectMCP

Name	Type	Description	Required	Default
type	OpenAI.ToolChoiceObjectType	Indicates that the model should use a built-in tool to generate a response..	Yes

OpenAI.ToolChoiceObjectCodeInterpreter

Name	Type	Description	Required	Default
type	enum	Possible values: code_interpreter	Yes

OpenAI.ToolChoiceObjectComputer

Name	Type	Description	Required	Default
type	enum	Possible values: computer_use_preview	Yes

OpenAI.ToolChoiceObjectFileSearch

Name	Type	Description	Required	Default
type	enum	Possible values: file_search	Yes

OpenAI.ToolChoiceObjectFunction

Use this option to force the model to call a specific function.

Name	Type	Description	Required	Default
name	string	The name of the function to call.	Yes
type	enum	For function calling, the type is always function. Possible values: function	Yes

OpenAI.ToolChoiceObjectImageGen

Name	Type	Description	Required	Default
type	enum	Possible values: image_generation	Yes

OpenAI.ToolChoiceObjectMCP

Use this option to force the model to call a specific tool on a remote MCP server.

Name	Type	Description	Required	Default
name	string	The name of the tool to call on the server.	No
server_label	string	The label of the MCP server to use.	Yes
type	enum	For MCP tools, the type is always mcp. Possible values: mcp	Yes

OpenAI.ToolChoiceObjectType

Indicates that the model should use a built-in tool to generate a response.

Property	Value
Description	Indicates that the model should use a built-in tool to generate a response.
Type	string
Values	file_search function computer_use_preview web_search_preview image_generation code_interpreter mcp

OpenAI.ToolChoiceObjectWebSearch

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
type	enum	Possible values: web_search_preview	Yes

OpenAI.ToolChoiceOptions

Controls which (if any) tool is called by the model.

none means the model will not call any tool and instead generates a message.

auto means the model can pick between generating a message or calling one or more tools.

required means the model must call one or more tools.

Property	Value
Description	Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools.
Type	string
Values	none auto required

OpenAI.ToolType

A tool that can be used to generate a response.

Property	Value
Description	A tool that can be used to generate a response.
Type	string
Values	file_search function computer_use_preview web_search_preview mcp code_interpreter image_generation local_shell

OpenAI.TopLogProb

The top log probability of a token.

Name	Type	Description	Required	Default
bytes	array		Yes
logprob	number		Yes
token	string		Yes

OpenAI.UpdateVectorStoreFileAttributesRequest

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	Yes

OpenAI.UpdateVectorStoreRequest

Name	Type	Description	Required	Default
expires_after	object	The expiration policy for a vector store.	No
└─ anchor	enum	Anchor timestamp after which the expiration policy applies. Supported anchors: last_active_at. Possible values: last_active_at	No
└─ days	integer	The number of days after the anchor time that the vector store will expire.	No
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	No
name	string	The name of the vector store.	No

OpenAI.ValidateGraderRequest

Name	Type	Description	Required	Default
grader	object	A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.	Yes
└─ calculate_output	string	A formula to calculate the output based on grader results.	No
└─ evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	No
└─ graders	object		No
└─ image_tag	string	The image tag to use for the python script.	No
└─ input	array	The input text. This may include template strings.	No
└─ model	string	The model to use for the evaluation.	No
└─ name	string	The name of the grader.	No
└─ operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	No
└─ range	array	The range of the score. Defaults to [0, 1].	No
└─ reference	string	The text being graded against.	No
└─ sampling_params		The sampling parameters for the model.	No
└─ source	string	The source code of the python script.	No
└─ type	enum	The object type, which is always multi. Possible values: multi	No

OpenAI.ValidateGraderResponse

Name	Type	Description	Required	Default
grader	object	A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.	No
└─ calculate_output	string	A formula to calculate the output based on grader results.	No
└─ evaluation_metric	enum	The evaluation metric to use. One of fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l. Possible values: fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, rouge_l	No
└─ graders	object		No
└─ image_tag	string	The image tag to use for the python script.	No
└─ input	array	The input text. This may include template strings.	No
└─ model	string	The model to use for the evaluation.	No
└─ name	string	The name of the grader.	No
└─ operation	enum	The string check operation to perform. One of eq, ne, like, or ilike. Possible values: eq, ne, like, ilike	No
└─ range	array	The range of the score. Defaults to [0, 1].	No
└─ reference	string	The text being graded against.	No
└─ sampling_params		The sampling parameters for the model.	No
└─ source	string	The source code of the python script.	No
└─ type	enum	The object type, which is always multi. Possible values: multi	No

OpenAI.VectorStoreExpirationAfter

The expiration policy for a vector store.

Name	Type	Description	Required	Default
anchor	enum	Anchor timestamp after which the expiration policy applies. Supported anchors: last_active_at. Possible values: last_active_at	Yes
days	integer	The number of days after the anchor time that the vector store will expire.	Yes

OpenAI.VectorStoreFileAttributes

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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.

Type: object

OpenAI.VectorStoreFileBatchObject

A batch of files attached to a vector store.

Name	Type	Description	Required	Default
created_at	integer	The Unix timestamp (in seconds) for when the vector store files batch was created.	Yes
file_counts	object		Yes
└─ cancelled	integer	The number of files that where cancelled.	No
└─ completed	integer	The number of files that have been processed.	No
└─ failed	integer	The number of files that have failed to process.	No
└─ in_progress	integer	The number of files that are currently being processed.	No
└─ total	integer	The total number of files.	No
id	string	The identifier, which can be referenced in API endpoints.	Yes
object	enum	The object type, which is always vector_store.file_batch. Possible values: vector_store.files_batch	Yes
status	enum	The status of the vector store files batch, which can be either in_progress, completed, cancelled or failed. Possible values: in_progress, completed, cancelled, failed	Yes
vector_store_id	string	The ID of the vector store that the File is attached to.	Yes

OpenAI.VectorStoreFileObject

A list of files attached to a vector store.

Name	Type	Description	Required	Default
attributes	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters, booleans, or numbers.	No
chunking_strategy	object		No
└─ type	enum	Possible values: static, other	No
created_at	integer	The Unix timestamp (in seconds) for when the vector store file was created.	Yes
id	string	The identifier, which can be referenced in API endpoints.	Yes
last_error	object	The last error associated with this vector store file. Will be null if there are no errors.	Yes
└─ code	enum	One of server_error or rate_limit_exceeded. Possible values: server_error, unsupported_file, invalid_file	No
└─ message	string	A human-readable description of the error.	No
object	enum	The object type, which is always vector_store.file. Possible values: vector_store.file	Yes
status	enum	The status of the vector store file, which can be either in_progress, completed, cancelled, or failed. The status completed indicates that the vector store file is ready for use. Possible values: in_progress, completed, cancelled, failed	Yes
usage_bytes	integer	The total vector store usage in bytes. Note that this may be different from the original file size.	Yes
vector_store_id	string	The ID of the vector store that the File is attached to.	Yes

OpenAI.VectorStoreObject

A vector store is a collection of processed files can be used by the file_search tool.

Name	Type	Description	Required	Default
created_at	integer	The Unix timestamp (in seconds) for when the vector store was created.	Yes
expires_after	OpenAI.VectorStoreExpirationAfter	The expiration policy for a vector store.	No
expires_at	integer	The Unix timestamp (in seconds) for when the vector store will expire.	No
file_counts	object		Yes
└─ cancelled	integer	The number of files that were cancelled.	No
└─ completed	integer	The number of files that have been successfully processed.	No
└─ failed	integer	The number of files that have failed to process.	No
└─ in_progress	integer	The number of files that are currently being processed.	No
└─ total	integer	The total number of files.	No
id	string	The identifier, which can be referenced in API endpoints.	Yes
last_active_at	integer	The Unix timestamp (in seconds) for when the vector store was last active.	Yes
metadata	object	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, and querying for objects via API or the dashboard. Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.	Yes
name	string	The name of the vector store.	Yes
object	enum	The object type, which is always vector_store. Possible values: vector_store	Yes
status	enum	The status of the vector store, which can be either expired, in_progress, or completed. A status of completed indicates that the vector store is ready for use. Possible values: expired, in_progress, completed	Yes
usage_bytes	integer	The total number of bytes used by the files in the vector store.	Yes

OpenAI.VoiceIdsShared

Property	Value
Type	string
Values	alloy ash ballad coral echo fable onyx nova sage shimmer verse

OpenAI.WebSearchAction

Discriminator for OpenAI.WebSearchAction

This component uses the property type to discriminate between different types:

Type Value	Schema
find	OpenAI.WebSearchActionFind
open_page	OpenAI.WebSearchActionOpenPage
search	OpenAI.WebSearchActionSearch

Name	Type	Description	Required	Default
type	OpenAI.WebSearchActionType		Yes

OpenAI.WebSearchActionFind

Action type “find”: Searches for a pattern within a loaded page.

Name	Type	Description	Required	Default
pattern	string	The pattern or text to search for within the page.	Yes
type	enum	The action type. Possible values: find	Yes
url	string	The URL of the page searched for the pattern.	Yes

OpenAI.WebSearchActionOpenPage

Action type “open_page” - Opens a specific URL from search results.

Name	Type	Description	Required	Default
type	enum	The action type. Possible values: open_page	Yes
url	string	The URL opened by the model.	Yes

OpenAI.WebSearchActionSearch

Action type “search” - Performs a web search query.

Name	Type	Description	Required	Default
query	string	The search query.	Yes
type	enum	The action type. Possible values: search	Yes

OpenAI.WebSearchActionType

Property	Value
Type	string
Values	search open_page find

OpenAI.WebSearchPreviewTool

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
search_context_size	enum	High level guidance for the amount of context window space to use for the search. One of low, medium, or high. medium is the default. Possible values: low, medium, high	No
type	enum	The type of the web search tool. One of web_search_preview or web_search_preview_2025_03_11. Possible values: web_search_preview	Yes
user_location	object		No
└─ type	OpenAI.LocationType		No

OpenAI.WebSearchToolCallItemParam

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
action	object		Yes
└─ type	OpenAI.WebSearchActionType		No
type	enum	Possible values: web_search_call	Yes

OpenAI.WebSearchToolCallItemResource

web_search is not yet available via Azure OpenAI.

Name	Type	Description	Required	Default
action	object		Yes
└─ type	OpenAI.WebSearchActionType		No
status	enum	The status of the web search tool call. Possible values: in_progress, searching, completed, failed	Yes
type	enum	Possible values: web_search_call	Yes

PineconeChatDataSource

Name	Type	Description	Required	Default
parameters	object	The parameter information to control the use of the Pinecone data source.	Yes
└─ allow_partial_result	boolean	If set to true, the system will allow partial search results to be used and the request will fail if all partial queries fail. If not specified or specified as false, the request will fail if any search query fails.	No	False
└─ authentication	object		No
└─ key	string		No
└─ type	enum	Possible values: api_key	No
└─ embedding_dependency	object	A representation of a data vectorization source usable as an embedding resource with a data source.	No
└─ type	AzureChatDataSourceVectorizationSourceType	The differentiating identifier for the concrete vectorization source.	No
└─ environment	string	The environment name to use with Pinecone.	No
└─ fields_mapping	object	Field mappings to apply to data used by the Pinecone data source. Note that content field mappings are required for Pinecone.	No
└─ content_fields	array		No
└─ content_fields_separator	string		No
└─ filepath_field	string		No
└─ title_field	string		No
└─ url_field	string		No
└─ in_scope	boolean	Whether queries should be restricted to use of the indexed data.	No
└─ include_contexts	array	The output context properties to include on the response. By default, citations and intent will be requested.	No	[‘citations’, ‘intent’]
└─ index_name	string	The name of the Pinecone database index to use.	No
└─ max_search_queries	integer	The maximum number of rewritten queries that should be sent to the search provider for a single user message. By default, the system will make an automatic determination.	No
└─ strictness	integer	The configured strictness of the search relevance filtering. Higher strictness will increase precision but lower recall of the answer.	No
└─ top_n_documents	integer	The configured number of documents to feature in the query.	No
type	enum	The discriminated type identifier, which is always ‘pinecone’. Possible values: pinecone	Yes

ResponseFormatJSONSchemaRequest

Name	Type	Description	Required	Default
json_schema	object	JSON Schema for the response format	Yes
type	enum	Type of response format Possible values: json_schema	Yes

ResponseModalities

Output types that you would like the model to generate. Most models are capable of generating text, which is the default:

["text"]

The gpt-4o-audio-preview model can also be used to generate audio. To request that this model generate both text and audio responses, you can use:

["text", "audio"]

Array of: string