Responses API

OpenAI's next-generation stateful endpoint, with built-in support for tools, web search, file search, and multi-step tasks — gradually replacing Chat Completions.

Create a Response

POST /v1/responses

RequestREQUEST

POSThttps://test.kailin.com.cn/v1/responses

API KeyStored only in your browser, never uploaded

Body (JSON)Non-streaming{ "model": "gpt-4o-mini", "input": "Describe the Responses API in one sentence" }

ResponseRESPONSE

Fill in the request and hit Send — the response shows up here

Request

curl https://test.kailin.com.cn/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-token" \
  -d '{
    "model": "gpt-4o-mini",
    "input": "Describe the Responses API in one sentence"
  }'

Request Body

Parameter	Type	Required	Description
model	string	Yes	Model name
input	string / array	Yes	Input text, or an array of messages/content parts
stream	boolean	No	Stream output (richer event types)
instructions	string	No	System-level instructions
previous_response_id	string	No	Pass the previous Response ID for multi-turn continuity without resending history
tools	array	No	Tools: function, web_search_preview, etc.
temperature / top_p	number	No	Sampling parameters
max_output_tokens	number	No	Max output tokens
metadata	object	No	Custom key-value pairs (up to 16)
store	boolean	No	Whether to store the result for previous_response_id reference

Example with an array input (including an image):

{
  "model": "gpt-4o-mini",
  "input": [
    {
      "role": "user",
      "content": [
        { "type": "input_text", "text": "Describe this image" },
        { "type": "input_image", "image_url": "https://example.com/cat.jpg" }
      ]
    }
  ]
}

Response

{
  "id": "resp_abc123",
  "object": "response",
  "created_at": 1730000000,
  "status": "completed",
  "model": "gpt-4o-mini",
  "output": [
    {
      "type": "message",
      "id": "msg_abc123",
      "role": "assistant",
      "status": "completed",
      "content": [
        {
          "type": "output_text",
          "text": "The Responses API is OpenAI's next-generation stateful endpoint…",
          "annotations": []
        }
      ]
    }
  ],
  "usage": {
    "input_tokens": 12,
    "output_tokens": 30,
    "total_tokens": 42
  }
}

To extract text, iterate output items with type: "message" and read content[].text; the SDK also offers the resp.output_text shortcut.

Retrieve a Response

GET /v1/responses/{response_id}

curl https://test.kailin.com.cn/v1/responses/resp_abc123 \
  -H "Authorization: Bearer sk-your-token"

Returns the full Response object. Responses created with store: false cannot be retrieved.

Compact Context

POST /v1/responses/compact

Compacts a long conversation into a summary — useful for long-running agents to reduce token usage on subsequent requests:

curl https://test.kailin.com.cn/v1/responses/compact \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-token" \
  -d '{
    "model": "gpt-4o-mini",
    "input": [ ...long message array... ]
  }'

Parameter	Type	Required	Description
model	string	Yes	Model used for compaction
input	array	Yes	Message array to compact
instructions	string	No	Custom instructions for the summary

Returns the compacted summary object, which can be placed in the input of the next request.

note

WebSocket access (wss://.../v1/responses with response.create events) depends on gateway WS passthrough support — REST + SSE remains the primary client integration path and is not covered here.