FunASR
Industrial-grade speech recognition toolkit: 170x realtime, 50+ languages, speaker diarization, emotion detection, streaming, and OpenAI-compatible API.
FunASR scores 7.6/10 (Strong) in Voice & Realtime.
Language Python License MIT Stars 18.8k★ Type heavyweight Self-host Yes Repo modelscope/FunASR Homepage https://github.com/modelscope/FunASR
Quality
Strong 7.6
Adoption 57
Activity 77
Maturity 99
Community 58
Capability 91
Show the math
Overall 7.6/10
value = 7.6/10
score = value
Adoption 6.1/10
value = 56.9/100
score = 1 + 9 * value / 100
Activity 7.9/10
value = 76.5/100
score = 1 + 9 * value / 100
Maturity 9.9/10
value = 98.6/100
score = 1 + 9 * value / 100
Community 6.2/10
value = 57.6/100
score = 1 + 9 * value / 100
Capability 9.2/10
value = 90.9/100
score = 1 + 9 * value / 100
Key metrics
18.8k★ GitHub stars
355/90d Recent commits
4 devs/90d Recent contributors
90.91% F1 VAD F1, FLEURS-VAD-102 (FunASR fsmn-vad) FunASR's VAD (fsmn-vad) scores F1 90.91% on FLEURS-VAD-102 in the independent FireRedVAD evaluation (high miss-resistance but 44% false-alarm rate). Source: FireRedVAD GitHub (https://github.com/FireRedTeam/FireRedVAD).
3.6y Project age
active Last commit
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
Labels
- Self-hosted
Self-hosted
You deploy and operate it yourself; there is no hosted option here.
- Heavyweight
Heavyweight
Resource-heavy to run (GPU, large memory, or multi-service stack).
asr streaming diarization multilingual industrial