+200K developers
Billions of audio hours processed
Drop-in API integration
Understand real-world conversation
Our Platform
Our background
1
2
3
4
5
6
7
8
9
10
11
12
# download pretrained pipeline from Huggingface
from pyannote.audio import Pipeline
pipeline = Pipeline.from_pretrained('pyannote/speaker-diarization-community-1', token="HUGGINGFACE_TOKEN")
# perform speaker diarization locally
output = pipeline('/path/to/audio.wav')
# enjoy state-of-the-art speaker diarization
for turn, speaker in output.speaker_diarization:
print(f"{speaker} speaks between t={turn.start}s and t={turn.end}s")
Open source roots
Community-1
Open-source, community-supported model. Widely adopted for research and development.
State-of-the-art
Precision-2
Higher accuracy, advanced controls, and enterprise-grade tooling for production teams.
11.5K
1B+
225K
Speaker Diarization
Identifies and labels each speaker in multi-participant audio.
Overlapping Speech Detection
Detects and attributes overlapping speech to the correct speakers.
STT Orchestration
Combines diarization and transcription for speaker-attributed transcripts.
Voiceprint
Identifies specific individuals across sessions using voice biometric signatures.
Confidence Scores
Identifies complex segments for quality control.
Speaker Separation (Soon)
Isolates overlapping speakers into individual audio streams.
curl