Sign in
Responses

Create response

Responses-shaped chat-completions compatibility adapter for compatible LLM and vision-language deployments. Does not implement full state, include, reasoning, hosted-tool, conversation-item, or background-job semantics.

post/responses

Request body

modelstring required
instructionsstring

System-level instruction text mapped into a chat system message.

streamboolean

When true, the response is a server-sent event stream.

max_output_tokensinteger
temperaturenumber
top_pnumber
reasoning_effort'none' | 'low' | 'medium' | 'high' | 'xhigh' | 'max'

Reasoning effort for this request, sent at the top level beside text. Which of these values a model honors, and which it rejects, is model specific and is stated on that model's page. On a model that cannot enforce a response format while it is reasoning, a request carrying text.format or response_format is refused with 400 hosted_parameter_not_supported and param response_format unless it also sends reasoning_effort none.

Example request

{
  "model": "deepseek-v4-flash"
}

Response

Response object or SSE stream if stream=true

idstring required
objectstring required
created_atinteger
modelstring required
output_textstring

Convenience text extracted from output content when available.

outputobject[]

Example response

{
  "id": "resp_abc123",
  "object": "response",
  "usage": {
    "cost": 0.00294,
    "runinfra": {
      "cost_microcents": 294000
    }
  }
}

Changes