Create a transcription task

Submit public audio URLs to create an asynchronous Fun-ASR speech-recognition task.

Submit one or more publicly accessible audio URLs for asynchronous speech recognition. The platform does not upload or store local audio files.

POST/v1/audio/asr/transcriptions

Submit one or more publicly accessible audio URLs. The platform does not upload or store local audio files. After the API returns a task ID, use the retrieval API to obtain the transcript download URLs.

A single request accepts up to 100 audio URLs and supports HTTP, HTTPS, and OSS URLs.

Authorization

bearerAuth

AuthorizationBearer <token>

Use Bearer your-api-key in the Authorization header.

In: header

Request Body

application/json

Fun-ASR asynchronous transcription request.

TypeScript Definitions

Use the request body type in TypeScript.

model?string

Fun-ASR model ID, such as fun-asr. Actual availability is determined by the model list returned for the current API key.

Length1 <= length

file_url*string

One publicly accessible audio URL.

Formaturi

file_urls?array<>

Multiple publicly accessible audio URLs.

Items1 <= items <= 100

input?

parameters?

Fun-ASR extension parameters. Other model-supported parameters may be supplied in addition to the documented fields.

Response Body

`application/json`

curl -X POST "https://example.com/v1/audio/asr/transcriptions" \  -H "Content-Type: application/json" \  -d '{    "model": "MODEL_ID",    "file_url": "https://example.com/audio.wav"  }'

{
  "task_id": "ff2e443d-976a-4d1c-b48d-72b92d8dfc7a",
  "status": "IN_PROGRESS",
  "model": "MODEL_ID"
}

Authorization

Request Body

Response Body

200application/json

401application/json

`application/json`

`application/json`