faster-whisper

Use it when

• transcribe long audio locally with efficient Whisper inference
• reduce transcription time and memory compared with the reference Whisper implementation

What it solves

Not the fit when

• speaker diarization requires an additional model or pipeline

Install

pip install faster-whisper

Invoke

from faster_whisper import WhisperModel; model = WhisperModel('large-v3'); segments, info = model.transcribe('audio.mp3')

Alternatives

No reviewed alternatives recorded yet.