Features
- Transcribes audio and video files, YouTube links and microphone input
- Runs locally on your device using Whisper models
- Translates audio and text
- Supports multiple engines including Whisper, Faster Whisper and the OpenAI Whisper API
- Separates speech before transcription to improve accuracy in noisy recordings
- Speaker recognition and word-level timestamps
- Exports to TXT, SRT and VTT
- Transcript viewer with search, playback speed control and keyboard shortcuts
- Watch Folder for automatic transcription of new files
- Command-line interface and plugin system with AI summarization
- GPU acceleration with CUDA, Apple Silicon and Vulkan
Pricing Details
Buzz is based on a free and open-source model and is distributed under the MIT License, with no subscriptions or paid plans listed in the official repository. The software offers audio transcription and subtitling capabilities for audio and video files, as well as real-time transcription from a microphone, speaker recognition, speech segmentation, and support for YouTube links, with the ability to export results in TXT, SRT, and VTT formats. It also supports advanced features such as Watch Folder for automatic file processing, a command-line interface, a plugin system, and on-device processing, along with processing acceleration using CUDA, Apple Silicon, and Vulkan, depending on the device and model used. Since the project is open source, there are currently no paid plans or announced subscription tiers, and it can be used and installed for free under the terms of the MIT license.