## Retrieve a model response

**get** `/api/v1/responses/{response_id}`

Retrieves a tenant-scoped stored Responses API response.

### Path Parameters

- `response_id: string`

### Header Parameters

- `"X-Organization-Id": optional string`

### Returns

- `Response object { id, object, created_at, 10 more }`

  - `id: string`

  - `object: "response"`

    - `"response"`

  - `created_at: optional number`

    Unix timestamp (in seconds) when the response was created.

  - `status: "completed" or "failed" or "in_progress" or 3 more`

    - `"completed"`

    - `"failed"`

    - `"in_progress"`

    - `"cancelled"`

    - `"queued"`

    - `"incomplete"`

  - `model: string`

    Public name of the model that generated the response.

  - `output: array of ResponseOutputItem`

    Items the model generated, in order.

    - `type: string`

      Item type, such as `message`, `reasoning`, `function_call`, `mcp_call` or `file_search_call`. Fields beyond the ones listed depend on the type.

    - `id: optional string`

      Identifier of the output item.

    - `status: optional string`

      Status of the item, such as `completed` or `in_progress`.

    - `role: optional "assistant"`

      Author of a `message` item.

      - `"assistant"`

    - `content: optional array of ResponseOutputContent`

      Content parts of a `message` item.

      - `type: string`

        Content part type: `output_text` for reply text, `refusal` for a refusal.

      - `text: optional string`

        The reply text, on an `output_text` part.

      - `refusal: optional string`

        The refusal message, on a `refusal` part.

      - `annotations: optional array of map[unknown]`

        Citations into `text`, such as file citations from file search.

    - `name: optional string`

      Name of the tool called, on `function_call`, `mcp_call` and `mcp_approval_request` items.

    - `arguments: optional string`

      JSON-encoded arguments for the tool, on `function_call`, `mcp_call` and `mcp_approval_request` items.

    - `call_id: optional string`

      Identifier to answer a `function_call` with: send it back as the `call_id` of a `function_call_output` input item.

    - `server_label: optional string`

      Label of the MCP server, on `mcp_call`, `mcp_approval_request` and `mcp_list_tools` items.

    - `approval_request_id: optional string`

      The approval request this `mcp_call` answers, if it needed approval.

    - `output: optional string`

      What the tool returned, on an `mcp_call` item.

    - `error: optional string`

      Why the tool call failed, on `mcp_call` and `mcp_list_tools` items.

    - `tools: optional array of map[unknown]`

      Tools the server offers, on an `mcp_list_tools` item.

    - `summary: optional array of map[unknown]`

      Summary parts of a `reasoning` item, each with `type` and `text`.

    - `encrypted_content: optional string`

      Opaque reasoning state to replay in a later request, on a `reasoning` item.

    - `queries: optional array of string`

      Search queries, on a `file_search_call` item.

    - `results: optional array of map[unknown]`

      Matched chunks, on a `file_search_call` item.

  - `output_text: string`

    The text of every `output_text` part in `output`, joined. Reasoning and tool calls are excluded.

  - `error: optional ResponseError`

    - `code: string`

    - `message: string`

  - `incomplete_details: optional ResponseIncompleteDetails`

    - `reason: optional string`

      Why the response is incomplete.

  - `usage: optional ResponseUsage`

    - `input_tokens: number`

    - `output_tokens: number`

    - `total_tokens: number`

    - `input_tokens_details: optional map[unknown]`

    - `output_tokens_details: optional map[unknown]`

  - `previous_response_id: optional string`

    The response this one continues, if any.

  - `store: optional boolean`

    Whether the response is stored for later retrieval.

  - `metadata: optional map[string]`

### Example

```http
curl https://api.prod.adaptionlabs.ai/api/v1/responses/$RESPONSE_ID \
    -H "Authorization: Bearer $ADAPTION_API_KEY"
```

#### Response

```json
{
  "id": "resp_123",
  "object": "response",
  "created_at": 1750000000,
  "status": "completed",
  "model": "qwen3.5-397b",
  "output": [
    {
      "type": "message",
      "id": "id",
      "status": "completed",
      "role": "assistant",
      "content": [
        {
          "type": "output_text",
          "text": "Once upon a time, a unicorn drifted to sleep on a cloud.",
          "refusal": "refusal",
          "annotations": [
            {
              "foo": "bar"
            }
          ]
        }
      ],
      "name": "get_weather",
      "arguments": "{\"city\":\"Paris\"}",
      "call_id": "call_123",
      "server_label": "docs",
      "approval_request_id": "approval_request_id",
      "output": "output",
      "error": "error",
      "tools": [
        {
          "foo": "bar"
        }
      ],
      "summary": [
        {
          "foo": "bar"
        }
      ],
      "encrypted_content": "encrypted_content",
      "queries": [
        "string"
      ],
      "results": [
        {
          "foo": "bar"
        }
      ]
    }
  ],
  "output_text": "Once upon a time, a unicorn drifted to sleep on a cloud.",
  "error": {
    "code": "server_error",
    "message": "The model provider could not complete the request."
  },
  "incomplete_details": {
    "reason": "max_output_tokens"
  },
  "usage": {
    "input_tokens": 36,
    "output_tokens": 87,
    "total_tokens": 123,
    "input_tokens_details": {
      "cached_tokens": "bar"
    },
    "output_tokens_details": {
      "reasoning_tokens": "bar"
    }
  },
  "previous_response_id": "previous_response_id",
  "store": true,
  "metadata": {
    "foo": "string"
  }
}
```