4 min read
How to Transcribe Audio or Video to Text Privately
Turn speech into text — and .srt/.vtt subtitles — in your browser with Whisper. No uploads, works offline after the first load.
Transcribing a voice note, meeting, podcast or video usually means uploading your recording to a cloud service. If the audio is sensitive — or you just don't want it on someone's server — you can transcribe it entirely on your own device instead.
Transcribe in 3 steps
- Open the Transcriber and drop in an audio or video file.
- Click "Transcribe". The first time, a small speech model (~40MB) downloads once, then it works offline. The audio is extracted and converted to text on your device.
- Copy the transcript, or download it as plain text (.txt) or subtitles (.srt / .vtt) with timestamps.
It uses OpenAI's Whisper model running in your browser via WebAssembly (and WebGPU when available) — your recording is never uploaded.
Try the TranscriberTranscribe audio or video to text on-device with Whisper — get a transcript plus .srt/.vtt subtitles, nothing uploaded.Tips for the best results
- It's tuned for clear English speech — accents, heavy noise or overlapping voices reduce accuracy.
- Shorter clips transcribe fastest; long files run for a while since it all happens on your device.
- Browsers with WebGPU (recent Chrome/Edge) are noticeably faster than the WASM fallback.
- Use the .srt or .vtt download to caption a video in any editor or player.