Voice Notes: Free Browser Dictation, Nothing Installed

Talk, get text. Uses the Web Speech API built into your browser.

Runs entirely in your browser. Nothing is uploaded, no account needed.

Idle
0 words

Your note is held in this tab only and is gone when you close it. Note that Chrome's implementation sends audio to Google's speech service for recognition, and Safari to Apple's — that request is the browser's, not this page's.

What this does

Typing out why a page was worth keeping takes long enough that most people skip it, and a link with no context is a link you will not recognize in March. Press record, say the sentence, edit the two words it got wrong. This is the same Web Speech API integration that Inspistor uses for the note on a saved link.

How to use it

  1. Press record and talk

    Interim results appear greyed out as you speak and are committed to the note when the recognizer is confident. Chrome, Edge and Safari support this; Firefox does not ship the API and the page tells you so rather than failing silently.

  2. Edit like a normal textarea

    The transcript is editable text, not a locked-off widget. Punctuation commands work in Chrome: say "period", "comma", "question mark", "new paragraph".

  3. Keep it

    Copy to clipboard, or download as .txt or .md. Notes are held in this tab only and are gone when you close it, which is either the feature or the caveat depending on what you came for.

Frequently asked questions

Does my audio get uploaded?

This page never touches your audio, but be precise about what the browser does: Chrome's implementation of the Web Speech API sends audio to Google's speech service for recognition, and Safari uses Apple's. That is the browser's network request, not ours, and it is the same path any web dictation feature takes. Nothing is stored by this page, and the text stays in the tab.

Which browsers work?

Chrome, Edge, Brave, Arc and Safari (including iOS Safari) support SpeechRecognition. Firefox does not implement it. On Chrome the recognizer stops itself after a pause, so this page restarts it automatically for as long as you leave recording on.

Can I dictate in another language?

Yes. Pick from the language list before recording; it sets the recognizer's lang, which changes both the acoustic model and the punctuation handling. Switching languages mid-note restarts the recognizer.

Why did it stop after 60 seconds?

Chrome ends a recognition session on silence and enforces its own timeouts. Continuous mode plus an auto-restart on the end event, which this page does, works around it. You will occasionally lose the last half-second across a restart boundary.

Built by Inspistor

We make a link library for people whose work spans more than one project: unlimited nesting, one-click Chrome capture, voice notes, and instant search.

Open the dashboard