FunASR
Industrial-grade speech recognition toolkit: 170x realtime, 50+ languages, speaker diarization, emotion detection, strea
Last Updated
2026-06-15
GitHub Stars
18,043
About FunASR
Our Verdict
A voice agent that is open-source and self-hostable backed by 18.0k GitHub stars. Industrial-grade speech recognition toolkit: 170x realtime, 50+ languages, speaker diarization, emotion detection, streaming, and OpenAI-compatible API..
Detailed Ratings
Pros & Cons
Pros
- Open-source — fully self-hostable and auditable
- Active community with 18.0k GitHub stars
- Designed for voice agent
Cons
- Requires technical setup; less hand-holding than managed alternatives
- Limited third-party reviews and benchmarks
Use Cases
Who Is It For?
Teams and individuals looking for voice and speech automation who value open-source flexibility and control.
Frequently Asked Questions
What is FunASR?
([简体中文](./README_zh.md)|English|[日本語](./README_ja.md)|[한국어](./README_ko.md))
How much does FunASR cost?
FunASR is open-source and free to self-host. See https://github.com/modelscope/FunASR for installation instructions.
Is FunASR open source?
Yes — FunASR is open source with 18.0k stars. The source code is on GitHub at https://github.com/modelscope/FunASR.
Who should use FunASR?
Teams and individuals looking for voice and speech automation who value open-source flexibility and control.
Related Agents
ElevenLabs
AI voice generation and text-to-speech platform
Deepgram
AI speech recognition and understanding platform
Leon
Leon is an open-source personal assistant who can live on your server
Retell AI
Conversational voice AI for businesses