Uses your browser's speech engine. On Chrome, this sends audio to Google's
speech API for the live preview only. The authoritative transcript below
still runs on our self-hosted Whisper — untick the box if you don't want
the preview.
Disables the browser's noise-suppression and echo-cancellation
(they filter out distant voices as "noise"), and boosts the mic
by 1.5×. Turn off if you're alone in a quiet room and want the
cleanest possible single-voice capture.
Auto-detect is disabled on live for now — it can trip a backend encoding bug on some languages. Pick yours here.
Start recording
Free — up to 20 minutes per session
Permission will be requested when you start. For tab audio, Chrome will show
a "Share this tab" dialog — tick Share tab audio. Your recording stays on our GPUs —
we don't train AI on it and we don't keep it longer than needed.
Warming up the mic…
Stop
First transcript appears ~15 s after you start.
Your transcript appears here. With live preview on, words show up as you speak; authoritative rows with timestamps appear every ~5–15 s.
You've used your 20 free minutes
Top up to keep going. PAYG credits never expire — $1 buys an hour.
See pricing →Something went wrong.
Works best in Chrome or Safari on desktop. Tab-audio capture (Zoom/Meet) requires Chrome's
share tab audio option — we'll add a system-audio capture button later.