Speech to text, on your device
Record or open an audio or video file and get an editable transcript. Your audio never leaves this browser.
- Transcribed on your device
- 99 languages
- Word, PDF, SRT, VTT
- Free to use
This recording
Saved to your account
- Source
- Length
- Windows
- Engine
- Checked when the model starts
- Language
- Not known until it is transcribed
This transcript came from your account; the audio stays wherever you keep it.
Changing a setting transcribes again and replaces your edits.
Long recordings run best on a computer: a phone is slower and can run out of memory.
Keeping stores this recording in this browser until you delete it. Anyone using this device can play it.
The time remaining appears after the first 30-second window.
Marked passages
Marks become an index at the front of the PDF, with the page and line each one landed on.
Subtitles No transcript yet
Document Numbered lines · a timestamp on every line · US Letter
is saved. To merge, split or compress it — Steady PDF runs in your browser too, and that file stays on your device as well.
Saving keeps the text, times, speaker labels and marks in your account. Your audio is not uploaded.
Signing in reloads the page, then saves this transcript. Open the recording again afterwards unless it is kept here.
Kept on this device
Nothing is kept on this device. A recording is kept only when you press “Keep on this device”.
These stay in this browser until you delete them, and anyone using this device can play them.
How it works
-
Record or open a file
A voice memo, an interview, a video. Nothing is sent anywhere.
-
Transcribed on your device
Whisper AI runs in your browser. The first run downloads the model, about 71 MB.
-
Check, correct, export
Play along, fix lines, name speakers, then export SRT, Word or PDF.
Where your data goes
Your device
The audio is never uploaded.
Your account (optional)
The text, times and labels you chose to save. Share a link or delete it.
This page, beside a cloud service
| Typical cloud transcription | Quiet Transcript | |
|---|---|---|
| Your audio | Uploaded and stored on their servers | Stays on your device |
| To start | An account first | Press Record or open a file |
| Free tier | Minutes per month | Unmetered: your device does the work |
| Documents | Text or subtitles | Numbered PDF, Word, Markdown, SRT, VTT |
| Speed | Fast on any device | Depends on your device; best on a computer |
What you get
- 00:00QWhere were you on the night of the fourth?
- 00:03AAt home, with my sister. We watched the match.
- 00:09QDid anyone else see you there?
1
00:00:00,000 --> 00:00:02,600
Q: Where were you on the night of the fourth?
2
00:00:03,100 --> 00:00:08,400
A: At home, with my sister.Questions
Is my audio uploaded?
No. The speech model runs in your browser, so the recording is read and transcribed on your device. You can check: open your browser’s network panel and watch while you transcribe.
What does an account add?
A library of transcripts you choose to save, and read-only share links that expire. Saving sends the text, times and labels; the audio still never leaves your device. You can transcribe without an account.
Does it work on my phone?
Yes, on iPhone and Android, in Safari or Chrome. A phone is slower than a computer and can run out of memory on a long recording, so use it for shorter ones.
Which languages, and how accurate is it?
Ninety-nine languages, worked out from the first 30 seconds; you can change it if the guess is wrong. English is strongest. The larger model is better outside English. It transcribes and never translates.
How long can a recording be?
Up to 2 GB on a computer. For a long file, transcribe the first five minutes to check the result, then run the whole recording.
Is it free?
Transcribing on your device is free. We are testing a Pro plan for cloud backup of recordings and server transcription; it is not open yet and nothing is charged.
Who this page is for, and what it will not do
The page was built for people with an obligation to whoever is on the tape. Think of a journalist with a source, a lawyer with a client, a clinician with a patient, an HR investigator with a complainant. What they share is a reason not to send a recording anywhere.
That is why the model is downloaded to your device and the audio is not uploaded to us. A recording is read, transcribed and corrected on the machine in front of you, and you decide what is saved.
Pick the job that matches yours
- Speech to text covers voice memos, interviews and lectures, including how the language is chosen and how long files are handled.
- Video to text covers recorded calls and clips, and what happens to a video that has no sound track.
- Subtitle generator covers SubRip and WebVTT files and the reading-speed checks behind them.
- Dictation covers the live draft that follows your voice while you talk.
- Voice recorder covers recording, keeping a take on this device and playing it back.
What it will not do
It will not translate. It will not tell speakers apart, so every speaker label is one you type. It will not hand you a certified or verbatim record, because the result is an automatic draft that you check against the recording.
It will not keep anything you did not ask it to keep. Reload the page and an unsaved transcript is gone, which is the cost of never holding your words on a server by default.
A good first session
Open a short recording you know well, press Transcribe and read the draft against the sound. In a few minutes that shows how your language, your microphone and your device behave, before you give it an hour of tape.
Check the language printed with the transcript, since a wrong guess looks like a normal transcript. Then export in the form your next step needs: subtitles for a video, a numbered PDF for citing, Word for editing.
The work is done by a speech model running inside your browser. The guide explains each step in order, from pressing Record to exporting a numbered PDF.
Advertisement — it pays for this, and it never sees your recording