FunAudioLLM/SenseVoice

Multilingual Voice Understanding Model

52
/ 100
Established

Combines ASR, emotion recognition, and audio event detection in a single non-autoregressive end-to-end model trained on 400k+ hours across 50+ languages. Achieves 15× faster inference than Whisper-Large while supporting timestamp generation via CTC alignment, with export options for ONNX and libtorch. Integrates with FunASR framework for streamlined deployment with multi-concurrent request handling and client SDKs (Python, C++, Java, C#).

7,691 stars.

No Package No Dependents
Maintenance 6 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

7,691

Forks

708

Language

Python

License

Last pushed

Dec 30, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/FunAudioLLM/SenseVoice"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.