Buzz

Buzz: offline Whisper transcription and subtitles

Buzz is a free, open-source desktop application for transcribing and translating audio on your own computer. It is powered by OpenAI’s Whisper ecosystem and is useful for creators, researchers, journalists, students, podcasters, and teams that need local speech-to-text without building a command-line pipeline from scratch.

What Buzz does

• Audio and video import: load recordings and media files, then export transcripts as TXT, SRT, VTT, and CSV.
• Offline transcription: run local models through Whisper, Whisper.cpp, Faster Whisper, or compatible Hugging Face models.
• OpenAI API option: use the hosted Whisper API when cloud transcription is acceptable or local hardware is too slow.
• Live microphone transcription: transcribe or translate from your computer microphone, with performance depending on hardware and model size.
• Subtitle production: create SRT and VTT captions for YouTube, courses, interviews, and video editing workflows.
• Advanced viewer: search transcripts, play audio, adjust speed, loop segments, and edit text inline.
• CLI workflows: automate transcription from scripts and batch jobs.

Best use cases

• Podcast and interview transcription.
• Lecture notes, research recordings, and meeting archives.
• Subtitle creation for videos and courses.
• Local transcription for privacy-sensitive recordings.
• Translation to English or same-language transcription across many supported languages.

Official media and screenshot

Limitations

Whisper transcription is resource-intensive. Real-time performance depends on CPU, GPU, model size, language, and audio quality. For long files or weak hardware, local transcription can take time; the OpenAI API option trades privacy and cost for speed and convenience.

Sources checked

• Official Buzz GitHub repository
• Official Buzz documentation
• Buzz Captions product site
• OpenAI speech-to-text documentation