Qwen-Audio-3.0-Asr-Flash

Model IntroductionVideo

Qwen-Audio-3.0-ASR-Flash is a high-performance speech recognition model from Alibaba's Tongyi Lab, engineered for accurate, context-aware transcription across diverse languages and dialects.

It is a non-real-time ASR model accessible through Alibaba Cloud Model Studio, designed to convert audio into structured, readable text. The model builds upon the Fun-ASR series with upgraded brand and capabilities, supporting transcription of audio files up to 5 minutes in length and offering context-awareness for improved accuracy.

The model supports 30 languages and 16 dialects across 8 major Chinese dialect groups, including Mandarin, Wu, Xiang, Gan, Hakka, Min, and Cantonese. It features enhanced industry-specific vocabulary recognition, achieving over 95% accuracy in medical scenarios. Its native text polishing capability automatically removes repetitions and filler words, while advanced punctuation prediction and text normalization convert numbers, dates, and amounts into standard written formats. Additionally, it supports both precompiled and on-the-fly hotwords—up to 50 high-priority P0 hotwords with over 90% recall—to further boost domain-specific term accuracy.

API request example

Endpoint /v1/video/generations. Replace <API_KEY> with a key from the console.

Async: poll `GET /v1/video/generations/{task_id}` with response `id`. Allowed size: 720p / 1080p.

curl https://api.ravenrouter.ai/v1/video/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <API_KEY>" \
  -d '{
  "model": "Qwen-Audio-3.0-Asr-Flash",
  "prompt": "a cute cat walking on grass",
  "duration": 5,
  "size": "720p"
}'

More models: API Reference → Examples by model

Qwen-Audio-3.0-Asr-Flash API — Video Generation Endpoint | RavenRouter AI