Transcribe audio & video to text
Real AI speech-to-text — OpenAI's Whisper model running inside your browser. Meetings, lectures, interviews, voice memos: get a transcript, SRT subtitles or VTT captions. No upload, no minutes limit, no subscription.
How it works
- Drop any audio or video file. MP3, WAV, M4A, MP4, MOV, WEBM and more — the audio track is extracted on your device.
- Pick a model and language. The fast model is built in and works offline; the higher-accuracy model downloads once (~74 MB). Language is auto-detected, or set it yourself.
- Transcribe. Watch live progress, then read the transcript and download it as TXT, SRT or VTT.
Frequently asked questions
Is my recording uploaded?
No. The Whisper AI model runs inside your browser — meetings, interviews and voice memos never leave your device. That's also why there's no minutes limit and no price.
Which languages are supported?
Whisper understands 90+ languages and auto-detects the spoken one. Setting it manually helps on short clips.
Can I get subtitles for a video?
Yes — drop the video itself. Download the result as SRT or WebVTT with timestamps, ready for YouTube or a video editor.
Why is it slower than online services?
Because the AI runs on your device instead of a server farm — that's what keeps your audio private and the tool free. On a laptop, expect roughly real-time or faster with the fast model; a browser with WebGPU speeds it up a lot.
How long can the recording be?
There's no hard limit — hour-long recordings work, they just take proportionally longer. The audio is processed in 30-second windows with a live progress bar.