Buzz is a desktop application for transcribing and translating audio locally with speech recognition models based on Whisper. It can process audio files, video files, and YouTube links without requiring cloud transcription. Live microphone transcription supports real-time captions and a presentation view for accessible events. Speech separation can improve results on noisy recordings, while speaker identification distinguishes voices within transcribed media. Multiple Whisper backends, Transformer models, and GPU acceleration options provide flexibility across different computers. Transcripts can be searched, reviewed alongside playback, exported in common subtitle formats, automated through a CLI, or extended with plugins.
Features
- Offline audio and video transcription
- Live microphone transcription
- Speaker identification and speech separation
- Multiple Whisper and Transformer backends
- TXT, SRT, and VTT transcript export
- CLI automation and plugin support