Buzz is a desktop application that transcribes and translates audio offline using OpenAI's Whisper model. It targets users who need private, on-device audio processing for transcription and translation tasks, offering a more secure and potentially faster alternative to cloud-based services.

How It Works

Buzz leverages OpenAI's Whisper, a powerful speech-to-text model, for its core transcription and translation capabilities. By running locally, it ensures data privacy and reduces reliance on external APIs, making it suitable for sensitive audio content. The application provides a user-friendly interface for managing audio files and accessing transcription results.

Quick Start & Requirements

• PyPI: pip install buzz-captions (requires ffmpeg)
• macOS: brew install --cask buzz or download .dmg from releases.
• Windows: Download and run .exe from releases (unsigned, requires manual override). winget install ChidiWilliams.Buzz
• Linux: Flatpak (flatpak install flathub io.github.chidiwilliams.Buzz) or Snap (sudo snap install buzz).
• GPU Support (Nvidia): Requires specific CUDA 12.1 PyTorch installation for the PyPI version.
• Documentation: Buzz Documentation

Highlighted Details

• Offline transcription and translation powered by OpenAI's Whisper.
• Cross-platform availability (macOS, Windows, Linux).
• GPU acceleration support for Nvidia cards via specific PyTorch installation.
• A native macOS version is available on the App Store with enhanced features.

Maintenance & Community

The project is maintained by Chidi Williams. Further community engagement details are not explicitly provided in the README.

Licensing & Compatibility

The README does not explicitly state a license. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The Windows .exe installer is unsigned, requiring users to bypass security warnings. GPU support requires a specific, manual installation of PyTorch with CUDA, which can be complex. The README does not detail the specific Whisper model version used or provide performance benchmarks.