Terms
Quiet Transcript creates an automatic draft, not a certified, sworn or verbatim transcript. Check the words against the recording before quoting, filing or relying on them.
What is offered
Quiet Transcript is provided as-is, with no warranty of accuracy, availability or fitness for a particular purpose. Transcribing on your device is free. The site may change or stop being available without notice, and nothing here is a service level.
How it is paid for
The site carries one advertisement, at the foot of the home page, from Google AdSense. No feature, format, length or quality depends on it, and blocking it changes nothing you can do here. Google is its own controller for what an advertisement request tells it about your visit. Your recording, your transcript and your corrections are never sent to Google or to an advertiser.
Pro, which is not open
We are testing interest in a paid Pro plan for cloud backup of recordings and server transcription. It is not open: pressing I’d pay for this records your interest and the price shown, and nothing is charged without asking you first. Prices shown are beta prices, may change before Pro opens, and do not include applicable tax. You can leave the list from your account page at any time.
Your account
An account is optional and needs a Google account. You are responsible for what you save and share, and for having the right to keep it: a transcript can contain other people’s words and sensitive personal information. This service offers no business associate agreement and is not for protected health information. We may suspend an account or disable a share link used to publish unlawful content or to harm others. You can delete your account, with everything it holds, at any time.
Share links
A share link lets anyone who has it read the transcript until it expires or you turn it off. Share only with people entitled to read it. Anyone may report a shared page through the link on it; we review reports and may disable the link.
What using it puts on your device
The first time you transcribe, your browser downloads about 71 MB of speech model and runtime from this site and keeps them. The work then happens on your own hardware, at your expense in time, battery and data.
Verify before you rely on it
Automatic transcription makes mistakes, most where the audio is worst — crosstalk, accents, distance from the microphone. The tool plays the recording beside the transcript and lets you correct any line so that checking is possible; the responsibility for doing it remains yours. It is not a substitute for a qualified transcriber where one is required, and nothing here is legal, medical or professional advice.
Three limits worth knowing
- Ninety-nine languages, and weaker in most of them. The model is
whisper-tiny, strongest in English; the largerwhisper-baseis offered on the tool page for an extra 80 MB. - The language is decided from the first thirty seconds. It is printed with the transcript and you can change it. The tool transcribes and never translates.
- No automatic speaker separation. Every speaker label is one you typed.
Your responsibility
You are responsible for having the right to record and transcribe the material you process, and for any consent or notification duty that applies where you are. The transcript is yours, and so is what you do with it.
What it is built from
The speech model is OpenAI’s whisper-tiny, licensed MIT, in the ONNX conversion published by onnx-community under Apache-2.0; whisper-base is the same licence in the same conversion. They run through Hugging Face’s transformers.js (Apache-2.0) on onnxruntime-web (MIT), served from this site and pinned by version and SHA-256.