Audio in.
Words out.
Drop a recording and read it back as text. The speech model runs inside this page, so your audio is never uploaded, and a long file costs no more than a short one.
Transcribe a file
Transcript
Private by design
The recording is decoded and transcribed on your device. There is no upload and no account.
One-time model download
The first run fetches the speech model (about 40 to 80 MB). Your browser keeps it, so later runs start straight away.
No hour cap
Unlocked, there is no limit on file length and no per-minute billing. Free try: the first 60 seconds of two files.
FAQ
What does the price buy?
£7.00 once, no subscription. It unlocks whole recordings of any length and any number of files, with Copy, TXT and SRT. The free try covers the first 60 seconds of 2 files.
Where does my audio go?
Nowhere. The file is decoded and transcribed inside this page. The only things fetched from the internet are the speech model (from Hugging Face, once, then cached) and the page's code libraries. Payment is handled by Stripe. The speech model downloads from Hugging Face. THATCHED keeps an anonymous count of page views and free uses, and anonymous error reports. Payment is handled by Stripe when you choose to pay.
Which browsers work?
Current Chrome, Edge, Firefox and Safari on desktop and phone. Very long files need plenty of free memory; on an older phone, split them first.
Refunds?
Digital goods: access starts straight away, so the 14-day cancellation right ends once you start using it. If something is broken, email business@thatched.ai.
Unlock the whole file
Pay once, transcribe recordings of any length, as many as you like. No subscription.