Sign in

v1

latestOpenAPI 3.1.02026-07-2219165126.1 KB
Responses

Create a response

Generate an OpenAI-compatible Responses API object from text or structured input items, with optional tools and streaming events.

post/v1/responses

Request body

modelstring required

Model identifier to route the request to.

instructionsstring nullable

A system (or developer) message inserted into the model's context.

max_output_tokensinteger nullable
temperaturenumber nullable
top_pnumber nullable
presence_penaltynumber nullable
frequency_penaltynumber nullable
parallel_tool_callsboolean nullable
truncation'auto' | 'disabled' nullable
streamboolean

Return a Responses API SSE event stream instead of a single JSON response.

Example request

{
  "input": "Explain the tradeoff between throughput and latency.",
  "instructions": "Respond in 2 bullet points.",
  "model": "gpt-4.1-mini"
}

Response

JSON response when stream=false, or Responses API Server-Sent Events when stream=true.

idstring required
object'response'
created_atinteger required
completed_atinteger nullable
statusstring required
incomplete_detailsobject nullable
modelstring required
previous_response_idstring nullable
instructionsstring nullable
truncation'auto' | 'disabled' nullable
parallel_tool_callsboolean nullable
top_pnumber nullable
presence_penaltynumber nullable
frequency_penaltynumber nullable
top_logprobsinteger nullable
temperaturenumber nullable
max_output_tokensinteger nullable
max_tool_callsinteger nullable
storeboolean nullable
backgroundboolean nullable
service_tierstring nullable
safety_identifierstring nullable
prompt_cache_keystring nullable

Example response

{
  "completed_at": 1743393601,
  "created_at": 1743393600,
  "id": "resp_123",
  "metadata": {},
  "model": "gpt-4.1-mini",
  "object": "response",
  "output": [
    {
      "content": [
        {
          "annotations": [],
          "logprobs": [],
          "text": "- Caching avoids repeated upstream work for identical or similar requests.\n- It reduces latency spikes and helps keep infrastructure costs predictable.",
          "type": "output_text"
        }
      ],
      "id": "msg_123",
      "role": "assistant",
      "status": "completed",
      "type": "message"
    }
  ],
  "status": "completed",
  "tools": [],
  "usage": {
    "input_tokens": 18,
    "input_tokens_details": {
      "cached_tokens": 0
    },
    "output_tokens": 31,
    "output_tokens_details": {
      "reasoning_tokens": 0
    },
    "total_tokens": 49
  }
}