Free Voice to Text

The tool uses your browser’s built-in speech recognition, part of the Web Speech API. When you press Start, the browser asks for your microphone, listens, and passes back the words it heard. The words it is still unsure about are shown in grey under the box; once settled, they are added to your text.

Choose the language you will speak in before you start. Recognition is much better when the language and accent match: English (UK), English (India) and English (Nigeria) each have their own option.

Everything else on this site runs only in your browser, and this site never receives your audio or your text here either. But speech recognition is done by the browser, and most browsers do it on their maker’s servers:

• Chrome sends the audio to Google’s speech service by default.
• Safari uses Apple’s dictation. Apple says audio is processed on the device when your settings show on-device dictation, and otherwise on Apple’s servers, without being stored unless you opted in to improving Siri and Dictation.
• Edge uses Microsoft’s online service; an on-device option is in preview.

Newer versions of Chrome can recognise some languages entirely on your device after a one-time download. Where your browser supports that, the tool shows Only recognise on this device; with it on, your audio does not leave your computer. For anything confidential, use that option or type instead.

Sources: MDN, Using the Web Speech API;Apple, Siri, Dictation and Privacy;Microsoft Edge documentation. Checked October 2026.

• Use a headset or sit close to the microphone, away from background noise.
• Speak in full phrases at a steady pace. Recognition uses the words around each word to choose between soundalikes such as “their” and “there”.
• Say punctuation as you go rather than adding it afterwards; it is quicker to fix a few words than a whole paragraph.
• When you are done, run the text through the text cleaner or the case converter if it needs tidying.