Retrieve a response - Alibaba Cloud Model Studio - Alibaba Cloud Documentation Center

Retrieve a completed model response by its Response ID.

China (Beijing)

The base_url for SDK calls is: https://dashscope.aliyuncs.com/compatible-mode/v1

HTTP request URL: GET https://dashscope.aliyuncs.com/compatible-mode/v1/responses/{response_id}

Path parameters

response_id string (required)

The Response ID to retrieve. The format is resp_xxx. You can get it from the response of the Create response API. Only Response IDs returned when store=true was set in the original creation request can be retrieved.

Python

import os
from openai import OpenAI

client = OpenAI(
    # If you have not configured an environment variable, replace the following line with: api_key="sk-xxx"
    api_key=os.getenv("DASHSCOPE_API_KEY"),
    base_url="https://dashscope.aliyuncs.com/compatible-mode/v1",
)

response = client.responses.retrieve("resp_xxx")
print(response)

Node.js

import OpenAI from "openai";

const openai = new OpenAI({
    // If you have not configured an environment variable, replace the following line with: apiKey: "sk-xxx"
    apiKey: process.env.DASHSCOPE_API_KEY,
    baseURL: "https://dashscope.aliyuncs.com/compatible-mode/v1"
});

async function main() {
    const response = await openai.responses.retrieve("resp_xxx");
    console.log(response);
}

main();

curl

curl https://dashscope.aliyuncs.com/compatible-mode/v1/responses/resp_xxx \
  -H "Authorization: Bearer $DASHSCOPE_API_KEY"

Response Returns the same Response object as the Create response API. The fields are described below:	{ "background": false, "completed_at": 1778676420, "created_at": 1778676418, "frequency_penalty": 0.0, "id": "resp_801bc2c4-93d9-910f-b35d-5274f5a737c1", "metadata": {}, "model": "qwen-plus", "object": "response", "output": [ { "content": [ { "annotations": [], "text": "Hello! Nice to meet you. How can I help you?", "type": "output_text" } ], "id": "msg_8c54756c-9b65-4a95-81d7-4276d91406db", "role": "assistant", "status": "completed", "type": "message" } ], "parallel_tool_calls": true, "presence_penalty": 0.0, "service_tier": "default", "status": "completed", "store": true, "temperature": 1.0, "tool_choice": "auto", "tools": [], "top_logprobs": 0, "top_p": 1.0, "usage": { "input_tokens": 45, "input_tokens_details": { "cached_tokens": 0 }, "output_tokens": 63, "output_tokens_details": { "reasoning_tokens": 0 }, "total_tokens": 108, "x_details": [ { "input_tokens": 45, "output_tokens": 63, "prompt_tokens_details": { "cached_tokens": 0 }, "total_tokens": 108, "x_billing_type": "response_api" } ] } }
id `string` The unique identifier for this response. The format is `resp_xxx`.
object `string` The object type. Always `response`.
status `string` The response status. Possible values: `completed` (completed), `failed` (failed), `in_progress` (in progress), `cancelled` (cancelled), `queued` (queued), `incomplete` (incomplete).
created_at `integer` The Unix timestamp (in seconds) when the response was created.
completed_at `integer` The Unix timestamp (in seconds) when the response finished generating. This field is `null` if the response is not complete.
error `object` An error object returned when the model fails to generate a response. This field is `null` on success.
model `string` The model ID used to generate the response.
output `array` An array of output items generated by the model. The type and order of elements depend on the model's response. Array element properties type `string` The output item type. Possible values: `message`: A message containing the model's final reply. `reasoning`: Reasoning output. Returned when `reasoning.effort` is set to a value other than `none` or when thinking mode is enabled. Reasoning tokens are counted in `output_tokens_details.reasoning_tokens` and billed accordingly. `function_call`: Function call output. Returned when using a custom `function` tool. You must handle the function call and return a result. `web_search_call`: Web search call output. Returned when using the `web_search` tool. `code_interpreter_call`: Code execution output. Returned when using the `code_interpreter` tool. `web_extractor_call`: Web extraction output. Returned when using the `web_extractor` tool. Must be used together with the `web_search` tool. `web_search_image_call`: Text-to-image search call output. Returned when using the `web_search_image` tool. Contains a list of images found. `image_search_call`: Image-to-image search call output. Returned when using the `image_search` tool. Contains a list of similar images found. `mcp_call`: MCP call output. Returned when using the `mcp` tool. Contains the result from the MCP service. `file_search_call`: Knowledge base search call output. Returned when using the `file_search` tool. Contains retrieval queries and results from the knowledge base. id `string` The unique identifier for the output item. Present for all output types. role `string` The message role. Always `assistant`. Present only when `type` is `message`. status `string` The output item status. Possible values: `completed` (completed), `in_progress` (in progress). Present when `type` is not `reasoning`. name `string` The tool or function name. Present when `type` is `function_call`, `web_search_image_call`, `image_search_call`, or `mcp_call`. For `web_search_image_call` and `image_search_call`, the value is fixed as `"web_search_image"` and `"image_search"`, respectively. For `mcp_call`, the value is the specific function name called in the MCP service (for example, `amap-maps-maps_geo`). arguments `string` The tool call arguments, formatted as a JSON string. Present when `type` is `function_call`, `web_search_image_call`, `image_search_call`, or `mcp_call`. Parse it using `JSON.parse()` before use. The contents vary by tool type: `web_search_image_call`: `{"queries": ["search keyword 1", "search keyword 2"]}`. The `queries` field contains a list of search keywords automatically generated by the model based on user input. `image_search_call`: `{"img_idx": 0, "bbox": [0, 0, 1000, 1000]}`. The `img_idx` field is the index of the input image (starting from 0). The `bbox` field contains the bounding box coordinates [x1, y1, x2, y2] of the search area, with values ranging from 0 to 1000. `function_call`: An argument object generated according to the user-defined function parameter schema. `mcp_call`: An argument object for the function called in the MCP service. call_id `string` The unique identifier for the function call. Present only when `type` is `function_call`. Use this ID to associate the request with the response when returning function call results. content `array` An array of message content. Present only when `type` is `message`. Array element properties type `string` The content type. Always `output_text`. text `string` The text content generated by the model. annotations `array` An array of text annotations. Usually an empty array. summary `array` An array of reasoning summaries. Present only when `type` is `reasoning`. Each element contains a `type` field (value `summary_text`) and a `text` field (the summary text). action `object` Search action information. Present only when `type` is `web_search_call`. Properties query `string` The search query keyword. type `string` The search type. Always `search`. sources `array` A list of search sources. Each element contains a `type` field and a `url` field. code `string` The code generated and executed by the model. Present only when `type` is `code_interpreter_call`. outputs `array` An array of code execution outputs. Present only when `type` is `code_interpreter_call`. Each element contains a `type` field (value `logs`) and a `logs` field (code execution logs). container_id `string` The code interpreter container identifier. Present only when `type` is `code_interpreter_call`. Use it to associate multiple code executions within the same session. goal `string` A description of the extraction goal, specifying what information to extract from the webpage. Present only when `type` is `web_extractor_call`. output `string` The output result of the tool call, formatted as a string. When `type` is `web_extractor_call`, this field contains a summary of the extracted webpage content. When `type` is `web_search_image_call` or `image_search_call`, this field is a JSON string containing an array of image search results. Each result includes a `title` (image title), `url` (image URL), and `index` (ordinal number) field. When `type` is `mcp_call`, this field is a JSON string result returned by the MCP service. urls `array` A list of URLs of the webpages extracted. Present only when `type` is `web_extractor_call`. server_label `string` The MCP service label. Present only when `type` is `mcp_call`. Identifies the MCP service used for this call. queries `array` A list of queries used for knowledge base retrieval. Present only when `type` is `file_search_call`. Each array element is a string representing a search query generated by the model. results `array` An array of knowledge base retrieval results. Present only when `type` is `file_search_call`. Array element properties file_id `string` The file ID of the matched document. filename `string` The filename of the matched document. score `float` The relevance score, ranging from 0 to 1. A higher value indicates greater relevance. text `string` A snippet of the matched document content.
usage `object` Token usage information for this request. Properties input_tokens `integer` The number of input tokens. output_tokens `integer` The number of output tokens generated by the model. total_tokens `integer` The total number of tokens used, equal to the sum of input_tokens and output_tokens. input_tokens_details `object` Fine-grained classification of input tokens. Properties cached_tokens `integer` The number of tokens that hit the cache. output_tokens_details `object` Fine-grained classification of output tokens. Properties reasoning_tokens `integer` The number of reasoning tokens. x_details `array` Billing details. Properties input_tokens `integer` The number of input tokens for this billing type. output_tokens `integer` The number of output tokens for this billing type. total_tokens `integer` The total number of tokens for this billing type. x_billing_type `string` Always `response_api`. prompt_tokens_details `object` Returned when Session caching is enabled. Contains a `cached_tokens` field (number of tokens that hit the cache). x_tools `object` Tool usage statistics. When built-in tools are used, this field contains the call count for each tool. Example: `{"web_search": {"count": 1}}`
tools `array` Echoes the full content of the `tools` parameter from the original Create response request. The structure matches the `tools` parameter in the request body. An empty array `[]` if no tools were used.
tool_choice `string` Echoes the value of the `tool_choice` parameter from the original Create response request. Possible values: `auto`, `none`, `required`.
parallel_tool_calls `boolean` Echoes the value of the `parallel_tool_calls` parameter from the original Create response request. Indicates whether the model is allowed to call multiple tools in parallel.
temperature `float` Echoes the value of the `temperature` parameter from the original Create response request. The sampling temperature controls the diversity of the model's output. Valid range: [0, 2). Returns the model's default value if not set.
top_p `float` Echoes the value of the `top_p` parameter from the original Create response request. The nucleus sampling probability threshold. Valid range: (0, 1.0]. Returns the model's default value if not set.
frequency_penalty `float` Echoes the value of the `frequency_penalty` parameter from the original Create response request. The frequency penalty coefficient. Positive values reduce the likelihood of repeated words.
presence_penalty `float` Echoes the value of the `presence_penalty` parameter from the original Create response request. The presence penalty coefficient. Positive values increase the likelihood of introducing new topics.
top_logprobs `integer` Echoes the value of the `top_logprobs` parameter from the original Create response request. The number of most likely tokens returned at each position. Returns `0` if not enabled.
store `boolean` Echoes the value of the `store` parameter from the original Create response request. `true` means the response is stored and can be referenced by `previous_response_id`. `false` means the response is not stored.
service_tier `string` The service tier. Always `default`.
background `boolean` Indicates whether the response was executed asynchronously in the background. Model Studio currently supports only synchronous calls, so this field is always `false`.
metadata `object` Echoes the content of the `metadata` parameter from the original Create response request. An object for attaching custom key-value pairs. An empty object `{}` if not set.

Error response

If the specified Response ID does not exist, the following error is returned:

{
    "error": {
        "message": "Response with id 'resp_xxx' not found.",
        "type": "InvalidParameter"
    }
}