modelscope/

FunASR

Industrial-grade speech recognition toolkit: 170x realtime, 50+ languages, speaker diarization, emotion detection, streaming, and OpenAI-compatible API.

Pythonasraudiochineseemotion-recognitionmcp-server+15 more

Stars

16.3k

+3 today+6 /wk+35 /mo

Forks

1.7k

Issues

Watchers

16.3k

Star History

Repository Info

LicenseMIT

CreatedNov 24, 2022

Last push2h ago

Homepagemodelscope.github.io/FunASR

Open on GitHub