Funasr-api: Production-ready Qwen3-ASR with 52-language support!

#12
by pengzhi27 - opened

Built a drop-in API service wrapping Qwen3-ASR (1.7B/0.6B) + FunASR models—fully compatible with OpenAI & Alibaba Cloud speech APIs.

What's inside:

  • Real-time streaming via WebSocket
  • Speaker diarization + word-level timestamps
  • Docker one-liner deploy

Perfect for anyone looking to run Qwen3-ASR in production without the infra headaches.

🔗 GitHub: https://github.com/Quantatirsk/funasr-api

Give it a spin and let me know what you think! 🙌

Here's a quick demo so you can see what it can do.

Sign up or log in to comment