Skip to content
buildbay.
Buzz logo

Buzz

chidiwilliams · Audio

GitHub stars

21,792
—Collecting 30d data

Stats updated

21,706 stars on Sep 27 to 21,792 stars on Oct 2, up 86.6 measured star snapshots

What is Buzz?

Buzz brings recorded media and microphone input into a transcript-review workflow. After local recognition, users can search and replay text, prepare timed subtitles, or automate repeated imports. Model backends and acceleration vary with the installation, and the current macOS release requires Apple Silicon.

Buzz is a desktop application for transcribing and translating audio offline with Whisper. It can process audio and video files, live microphone input, and export transcripts as TXT, SRT, or VTT.

Core capabilities

  • Transcribe or translate local audio, video, microphone input, and YouTube links
  • Run Whisper-family backends with Apple Silicon, CUDA, or Vulkan acceleration
  • Review, search, and export transcripts as TXT, SRT, or VTT

A local transcription pipeline

Buzz offers Whisper-family backends with Apple Silicon, CUDA, and Whisper.cpp/Vulkan acceleration options. Speech separation and speaker identification can help with difficult recordings. Availability depends on the selected backend, hardware, and release; current macOS builds require Apple Silicon, while 1.4.5 was the last Intel-compatible version.

  • offline
  • Whisper
  • CUDA
  • Vulkan

From capture to reusable text

The same application handles imported media, YouTube links, watched folders, and live microphone capture. Its transcript viewer joins search with playback and speed controls, then exports plain text or timed subtitle formats; CLI and plugin interfaces extend the process beyond one-off desktop jobs.

  • live audio
  • subtitles
  • CLI
  • plugins

Where it fits

Use cases

  1. 01

    Create subtitles from recorded media

    Import an audio or video file, choose a local model and language settings, inspect the transcript alongside playback, and export timed SRT or VTT captions for editing or publication.

  2. 02

    Provide live transcription at an event

    Capture microphone audio in real time and show the dedicated presentation window so spoken material remains visible during a meeting, lecture, or accessibility-focused event.

  3. 03

    Automate a recurring transcription queue

    Use a watched folder or the command-line interface to process incoming recordings repeatedly. Optional plugins can add follow-on work; assess a plugin's services and data handling separately from local transcription.