Dictating
Choose a language
Pick English (UK) or one of the other listed languages, including Welsh, French, German, Spanish, Italian, Polish, Portuguese and Dutch. It cannot be changed while listening.
Allow the microphone
Press Start listening and accept your browser's microphone prompt. Words appear in grey while they are being recognised and settle into the transcript when final.
Punctuate by voice
Leave the punctuation box ticked and say comma, full stop, question mark, exclamation mark, colon, new line or new paragraph. Sentences get a capital letter automatically.
Keep the text
Press Stop listening, then Copy text or Download .txt. A running word count sits beside the buttons; Clear starts again.
Example: call notes straight after hanging up
A sales rep finishes a call and, before the details fade, dictates a short note for the CRM. Spoken punctuation turns the words into a tidy entry with a line break where they said "new line".
The recogniser's own capitalisation varies, so names like Harbour Plumbing may need a quick fix before pasting, but the structure arrives intact.
"called harbour plumbing comma spoke to the office manager full stop new line quote sent on Monday comma follow up Thursday question mark"
Called harbour plumbing, spoke to the office manager. Quote sent on Monday, follow up Thursday?
Questions people ask
Which browsers support this speech to text tool?
Chrome and Edge on desktop and Android offer the speech recognition the page relies on. Firefox does not, and support elsewhere varies. If your browser lacks it, the page says so; your phone's keyboard dictation button is a good alternative, and you can paste the result into a document.
How accurate is browser dictation?
For clear speech in a quiet room with a decent microphone it is usually good, though names, trade terms and postcodes are often misheard. Speak in phrases rather than single words, check the transcript before using it, and correct names by hand.
Can it transcribe an audio file or a recorded call?
No. It listens to a live microphone only. For a recording you would need a transcription service that accepts files; if it requires MP3, the M4A to MP3 or WAV to MP3 tools can prepare the file first.
Where your data goes
The page asks for microphone access and passes the sound to your browser's speech recognition. In Chrome that audio goes to the browser maker's speech service, not to us; the transcript is shown and kept only in this tab.
Checked against the tool on