How It Works

Drop in a recording, watch the transcript appear line by line, fix anything that's off, and export captions — all without your audio ever leaving your computer.

1

Drop Your Audio

Drag in one or several recordings — interviews, lectures, meetings, voice memos. Nothing uploads; files are read directly from your computer.

2

Watch It Transcribe Live

Words appear within seconds and are copyable right away — the app skips silence automatically and stays smooth even on very long files.

3

Fix Anything, Instantly

Click any line to correct it. Timestamps stay word-accurate, so captions re-flow instantly around your edits.

4

Export Captions or Text

Download SRT or VTT for YouTube, Premiere, and DaVinci Resolve, plain text for notes, or copy straight to your clipboard.

Privacy Guarantee

  • All transcription happens locally in your browser — zero data leaves your machine
  • Network is only used to download the speech recognition engine on first use (then cached locally)
  • Disconnect your internet after that — the app works entirely offline

What You Can Do

  • Formats: MP3, WAV, M4A, OGG, FLAC
  • Timestamps: word-level accuracy, not just paragraph guesses
  • Length: hours-long recordings, streamed rather than loaded whole
  • Export: SRT, VTT, TXT, or clipboard

Frequently Asked Questions

Is my audio uploaded anywhere? +
No. Speech recognition runs entirely on your device, inside your browser. Your files never leave your machine. You can verify this yourself: disconnect your internet after the one-time model download and the app keeps working.
Which audio formats are supported? +
MP3, WAV, M4A, OGG, and FLAC. Support for video files is planned.
How accurate is it? +
Accuracy depends on audio quality and the mode you choose. Standard mode runs everywhere and is very accurate for clear speech; Maximum mode uses your graphics card for a noticeably more accurate result, rivaling paid transcription services.
Why is there a one-time download? +
The speech recognition engine has to run on your device for the privacy guarantee to hold — that engine downloads once and is cached by your browser. Every visit after that starts instantly and works fully offline.
Does it work on long recordings? +
Yes. Hours-long recordings are fully supported — audio is streamed and processed in short windows rather than loaded into memory all at once, and you see the transcript appear as it's produced.
Can I edit the transcript and export subtitles? +
Yes. Click any line to correct it, then export as SRT or VTT for video captions, plain text, or copy straight to your clipboard.