BrowserTools

Speech to Text

Live dictation straight into text

By Guga — BrowserTools maintainer, Tbilisi, Georgia Updated 2026-09-09
Free Private No uploads

The page sends nothing itself, but the recogniser does: Chrome and Edge ship a cloud speech service that receives your microphone audio, while Safari's is on-device. Check your browser's documentation before dictating anything sensitive.

What is Speech to Text?

Words appear as you speak, in grey while the engine is still deciding and in black once it commits to them, and the committed text accumulates in a panel you can copy. It recognises continuous speech with interim results enabled, so you can talk for minutes without pausing for it. The language is hard-coded to en-US — there is no selector, and no attempt to detect or switch languages. This is dictation, not transcription: there is no way to feed it an existing recording, no timestamps, no speaker labels and no punctuation commands beyond what the engine adds on its own. The transcript is text and text only; nothing is written to a file unless you copy it somewhere yourself.

Dictate with your microphone and watch words appear as text in real time, ready to copy anywhere. Uses your browser's built-in speech recognition — no account needed.

Private Speech to Text In Your Browser

Speech to Text is processed locally with client-side JavaScript. Your files never touch a server, and there is no queue to wait in.

Step-by-Step

How to use Speech to Text

1

Press Start dictating and allow microphone access when the browser prompts.

2

Speak normally; a grey preview shows the current guess and the text turns solid once each phrase is finalised.

3

Press Stop listening when you are done, then use the copy button beside the word count to take the transcript with you.

4

Press Clear transcript to reset the panel for the next recording.

Key Capabilities
  • Runs continuous recognition with interim results, so committed words accumulate without pauses.
  • Shows a live word count next to the copy control and collapses repeated whitespace as text is appended.
  • Separates tentative text visually from final text, so you can see what is still being revised.
  • Uses the browser's built-in recognition with no engine download and no account.
Common Use Cases
  • A journalist drafting quotes at the end of an interview without reaching for a keyboard.
  • A student writing up lecture notes hands-free, reading them back afterwards and fixing the words the engine got wrong.
  • An engineer dictating a bug description into an issue tracker while reproducing the problem.
Clear Answers

Frequently Asked Questions

Does my microphone audio leave the machine?
This page does not transmit anything, but the recogniser does: in Chrome and Edge the built-in recogniser runs on the browser vendor's cloud speech service, which receives your audio. Safari's is on-device. If the material is sensitive, use Safari or check your browser's documentation for the current arrangement.
Can I dictate something other than English?
No. Recognition is fixed to en-US in this page, with no language control exposed. A foreign accent is usually handled, but a different language is not, and the page offers nothing to change that.
Why did it stop on its own mid-sentence?
That is normal behaviour for the continuous mode when a pause is detected — the session ends, the listening indicator clears, and you press Start dictating again. The transcript survives, so you can carry on from where it left off rather than starting over.
Can it transcribe a recording I already have?
No. Recognition only ever runs on the live microphone stream; there is no file input and no way to point it at a stored recording. Transcribing an existing file needs a different tool that can process media rather than capture it.