Skip to main content
POST
/
vendors
/
openai
/
v1
/
responses
Create model response
curl --request POST \
  --url https://api.mulerun.com/vendors/openai/v1/responses \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "gpt-5.2",
  "input": "What is the capital of France?"
}
'
{
  "id": "resp_67cb32528d6881909eb2859a55e18a85",
  "object": "response",
  "created_at": 1741369938,
  "model": "gpt-5.2",
  "status": "completed",
  "output": [
    {
      "id": "msg_67cb3252cfac8190865744873aada798",
      "type": "message",
      "role": "assistant",
      "content": [
        {
          "type": "output_text",
          "text": "The capital of France is Paris."
        }
      ]
    }
  ],
  "output_text": "The capital of France is Paris.",
  "usage": {
    "prompt_tokens": 33,
    "completion_tokens": 8,
    "total_tokens": 41,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "audio_tokens": 0,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0
    }
  }
}
This API is compatible with OpenAI’s Responses API format.For more details, please refer to OpenAI’s official documentation.

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json
​
model
string
required

Model deployment to use for response creation.

Example:

"gpt-5.2"

​
input

Input content for the model. Can be a simple text string or an array of structured message objects.

Example:

"Hello! How are you?"

​
instructions
string | null

System/developer message inserted into the model context to guide behavior, tone, and goals.

​
previous_response_id
string | null

ID of a previous response to use for multi-turn conversations and context continuation.

​
prompt
object

Reference to a prompt template with variables.

​
max_output_tokens
integer | null

Upper bound for the number of tokens that can be generated (includes reasoning tokens).

Required range: x >= 1
​
max_tool_calls
integer | null

Maximum total number of built-in tool calls allowed in the response.

​
temperature
number | null

Sampling temperature between 0 and 2. Higher values make output more random, lower values make it more focused and deterministic.

Required range: 0 <= x <= 2
Example:

0.7

​
top_p
number | null

Nucleus sampling parameter. Model considers tokens with top_p probability mass. For example, 0.1 means only tokens comprising the top 10% probability mass are considered.

Required range: 0 <= x <= 1
Example:

1

​
top_logprobs
integer | null

Number of most likely tokens to return at each token position (0-20).

Required range: 0 <= x <= 20
​
parallel_tool_calls
boolean | null

Allow the model to run tool calls in parallel during tool use.

​
stream
boolean | null

If true, the response will be streamed back using server-sent events.

​
background
boolean | null

If true, the response will run in the background. Use GET /v1/responses/{response_id} to poll for completion.

​
store
boolean | null

Whether to store the response for later retrieval. Stored responses can be accessed via the retrieve endpoint.

​
text
object

Configuration for text response format. Supports plain text or structured JSON data (Structured Outputs).

​
tools
object[] | null

List of tools the model may call. Can include built-in tools (code_interpreter, file_search, web_search, image_generation) or custom functions.

Tools the model may call while generating a response. Supports built-in tools (file_search, web_search, code_interpreter, image_generation), MCP Tools, and custom function calls.

​
tool_choice

How the model should select which tool (or tools) to use when generating a response.

Available options:
none,
auto,
required
​
reasoning
object

Configuration for reasoning models only (e.g., o-series models).

​
truncation
enum<string> | null

Truncation strategy for handling context window limits.

Available options:
auto,
disabled
​
metadata
object

Set of 16 key-value pairs for storing additional information. Keys max 64 characters, values max 512 characters.

​
user
string | null

Unique identifier representing the end-user, for abuse monitoring and detection.

​
include
string[] | null

Additional output data to include in the response. Examples: "code_interpreter_call.outputs", "file_search_call.results"

Response

200 - application/json

OK

Response object returned after creating a model response.

​
id
string
required

Unique identifier for the response

Example:

"resp_67cb32528d6881909eb2859a55e18a85"

​
object
enum<string>
required

Object type identifier

Available options:
response
​
created_at
number
required

Unix timestamp of when the response was created

Example:

1741369938

​
model
string
required

The model that was used to generate the response

​
status
enum<string>
required

The current status of the response

Available options:
completed,
incomplete,
in_progress,
failed
​
output
(Message output · object | Reasoning output · object | Function call output · object)[]

Array of output items from the model

​
output_text
string | null

Concatenated text from all text outputs in the response

​
usage
object

Token usage statistics for the response

​
error
object

Error information if the response failed