Jimeng Format

Jimeng's (即梦) official API-compatible format, using the Action query parameter to distinguish "submit task" from "get result" operations.

Different path prefix

The Jimeng format uses the /jimeng/ path prefix — without the /v1 prefix used elsewhere in this documentation. Authentication is still Authorization: Bearer <YOUR_API_KEY>.

Endpoint

POST /jimeng/

Query parameters

Parameter	Type	Required	Description
Action	string	Yes	CVSync2AsyncSubmitTask (submit) or CVSync2AsyncGetResult (get result)
Version	string	Yes	API version

Request body (submit task)

Parameter	Type	Required	Description
req_key	string	Yes	Request type identifier
prompt	string	No	Text description
binary_data_base64	array	No	Base64-encoded image data (for image-to-video)

Submit a video generation task

curl "https://ai.youqi.tech/jimeng/?Action=CVSync2AsyncSubmitTask&Version=2022-08-31" \
  -H "Authorization: Bearer sk-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "req_key": "jimeng_vgfm_t2v",
    "prompt": "An astronaut stands up and walks away"
  }'

Response example

{
  "code": 10000,
  "message": "Success",
  "data": { "task_id": "abcd1234efgh" }
}

Get task result

curl "https://ai.youqi.tech/jimeng/?Action=CVSync2AsyncGetResult&Version=2022-08-31" \
  -H "Authorization: Bearer sk-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "req_key": "jimeng_vgfm_t2v",
    "task_id": "abcd1234efgh"
  }'

TIP

req_key, Version, and other parameters must match Jimeng's official documentation — different task types (text-to-video / image-to-video) use different req_key values.