Buzz is a free, open-source desktop app that transcribes and translates audio offline, powered by OpenAI’s Whisper models. It has earned more than 22,000 stars on GitHub, is actively maintained with updates in October 2026, and is released under the MIT license.
Buzz brings Whisper’s speech recognition to people who do not want to touch a command line. Install the app, drop in an audio or video file, or record straight from your microphone, and Buzz produces a transcript on your own computer with no audio ever uploaded to a server. It supports multiple Whisper implementations under the hood, including faster backends that take advantage of modern hardware, and it can translate speech into English as it transcribes. Because everything runs locally, it suits confidential interviews, unreleased media, and any workflow where recordings cannot leave the building.
Buzz is designed for everyday transcription work rather than demos. Files can be queued and processed in batches, transcripts are presented as timestamped segments you can scan and search, and finished text can be exported for editing, captioning, or archiving. Live recording transcription helps with lectures and meetings as they happen, while file-based transcription suits recorded interviews and footage. Because Whisper supports many languages, the same workflow handles multilingual recordings, and the translation mode can turn foreign-language speech into English text. It has become a go-to free option for people who previously paid per minute for cloud transcription services.
Buzz is completely free and open source under the MIT license, with no accounts, subscriptions, or per-minute transcription fees.
Buzz is perfect for journalists, podcasters, researchers, students, and legal or medical professionals who need accurate transcripts and subtitles without paying for cloud transcription or exposing sensitive audio.