pyannote / pyannote-audio

git clone https://github.com/pyannote/pyannote-audio.git 4.0.7 · 3mo ago

Python toolkit for speaker diarization built on PyTorch, with pretrained pipelines and models that can be fine-tuned on your own data.

Why you might care

Gives you pretrained speaker diarization pipelines (community-1 open-source, plus a premium precision-2 option via pyannoteAI) and building blocks for speech activity, speaker change, overlapped speech detection, and speaker embeddings. Useful when you need to answer 'who spoke when' on meeting or call audio and want to fine-tune on your own data; multi-GPU training is supported via pytorch-lightning.

Alternatives

WhisperXSpeechBrainNVIDIA NeMoKaldi

stars, 24h · #4,239

+31

stars, 7 days · #6,293

+~120

stars, 30 days · #6,159

11k

★ total · +0% this week

1.1k

forks · 32 open issues

Star history (as recorded here)

2026-10-20 · 10,674 ★ 2026-10-24 · 10,690 ★

LanguageJupyter Notebook

LicenseMIT

Created2016-03-07 (10y ago)

Last push1d ago

Watchers83

OwnerOrganization

Seen on GitHub Trending

2026-10-23 — #14 weekly in jupyter-notebook · +31 ★
2026-10-21 — #19 weekly in jupyter-notebook · +40 ★
2026-10-20 — #19 weekly in jupyter-notebook · +42 ★

GitHub topics

overlapped-speech-detection pretrained-models pytorch speaker-change-detection speaker-diarization speaker-embedding speaker-recognition speaker-verification speech-activity-detection speech-processing voice-activity-detection

README

Loading README…

Blurb generated by claude-haiku-5-5 1d ago. Star deltas marked ~ are extrapolated from partial history.