Voice to Text
Press record and talk. The text appears as you speak and stays editable — say "full stop" or "comma" to punctuate.
Where your voice actually goes
This is the one place on the site worth reading carefully. Browser speech recognition uses the Web Speech API, and in Chrome and Edge that means your audio is sent to Google's or Microsoft's servers for processing — it is not done on your device. Safari on recent Apple systems can process some dictation locally. Firefox does not support the API at all. So this tool is genuinely free and works well, but it is not private in the way the rest of this site is. Do not dictate anything confidential.
Punctuation commands
| Say | Get |
|---|---|
| "full stop" or "period" | . |
| "comma" | , |
| "question mark" | ? |
| "exclamation mark" | ! |
| "colon" / "semicolon" | : ; |
| "new line" | A line break |
| "new paragraph" | A blank line |
| "dash" | — |
Getting better accuracy
- Speak at a normal pace. Slowing down and over-enunciating makes recognition worse, not better — the models are trained on natural speech.
- Reduce background noise. Music and other voices confuse it far more than volume does.
- Get closer to the microphone, but not so close that you produce plosives.
- Pick the right regional variant. English (UK) versus English (US) meaningfully changes accuracy on accents and vocabulary.
- Dictate in sentences. Longer stretches give the model context, and context is where most of its accuracy comes from.
- Expect to edit. Names, technical terms and homophones will need fixing — the box is editable for that reason.
How to use the Voice to Text
- Pick your language and press Start recording.
- Allow microphone access when your browser asks.
- Speak normally. Say "full stop", "comma" or "new paragraph" to punctuate.
- Edit the text directly in the box, then copy it.
Frequently asked questions
Is my voice sent to a server?
In Chrome and Edge, yes — audio goes to Google or Microsoft for processing. Safari on recent Apple systems can handle some dictation on-device. This is a property of the browser API, not of this page, and it is why you should not dictate anything confidential.
Why does it not work in Firefox?
Firefox does not implement the Web Speech recognition API. Use Chrome, Edge or Safari.
Why does it stop after a while?
Browsers end a recognition session after a pause. This page restarts it automatically while recording is on, so you should not notice — though very long sessions may occasionally need restarting manually.
Can I dictate in other languages?
Yes, fifteen are listed. Recognition quality varies considerably by language and accent.
How do I add punctuation?
Say it. "Full stop", "comma", "question mark" and the others in the table above are converted automatically.
Does it work on a phone?
Yes, in Chrome on Android and Safari on iOS. On a phone, your keyboard’s built-in dictation button often works just as well.
Related tools
Last reviewed September 4, 2026. Results are estimates for general information only.