Qwen ASR · Video Generation

Fast Recognition (Text)NEW

📌 What is this：Same submit + poll-by-taskId flow, but on succeeded the resultUrl is the recognized text directly (not a file URL).

💡 When to use：Use for short audio when you want the text quickly and need no timestamps.

Endpoints

Usage	Method	Path
Submit task	POST	/api/cqt/generator/qwen
Query Result	GET	/api/cqt/info/qwen?id={taskId}

⏱ Recommended polling interval：All 4 models return a taskId on submit; poll by taskId until status = succeeded / failed (every 5-10s). qwen3-asr-flash is fast and usually reaches succeeded quickly.

Request Parameters

Param	Type	Required	Description
model	string	Required	qwen3-asr-flash
audioUrl	string	Required	Input audio file URL (replaces prompt; ASR has no prompt).
duration	integer	Required	⚠️ Input audio duration in seconds — drives billing. Default 5; always pass the real length or billing is badly underestimated.
webhookUrl	string	—	Completion webhook URL (optional).

Request Example

curl -X POST https://api.cqtai.com/api/cqt/generator/qwen \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{ "model": "qwen3-asr-flash", "audioUrl": "https://cdn.example.com/a.wav", "duration": 10 }'
# -> { "code":200, "data":"<taskId>" }

Response Example

{
  "code": 200,
  "data": {
    "id": "<taskId>",
    "status": "succeeded",
    "resultUrl": "识别出的文本内容…"
  }
}

Billing & Credits

0.022 credits/sec (pre-charged by duration) (x your user rate). Auto-refunded on failure.