Free Speech to Text
Dictate in 80+ languages and accents with live results, spoken punctuation and an editable transcript you can copy or save as a .txt file. Free.
Loading tool…
How to type with your voice using live speech to text
Live dictation is useful for drafting emails, notes, essays and messages faster than typing, or when typing is uncomfortable. It is the reverse of the Text to Speech tool, which reads text aloud. Recognition is provided by your browser’s speech service, so results and language support depend on the browser you use.
- Open the page in Chrome, Edge or Safari. Firefox does not offer speech recognition to websites, and the page explains this if it is missing.
- Choose the language and regional accent you will speak, for example English (India) or Spanish (Mexico).
- Press Start dictation and allow microphone access. Grey italic text shows words still being recognised.
- In English, optionally say punctuation words such as comma, period, question mark or new paragraph.
- Correct any mistakes directly in the transcript box, then copy the text or download it as a .txt file.
When live dictation helps
Most people speak about three times faster than they type, so dictation is a quick way to get a first draft out of your head: an email reply, meeting notes, a journal entry, a story outline or a long message. It also helps people who find typing slow or painful, and people writing in a language whose keyboard layout they rarely use.
Because the transcript is editable, the usual workflow is to speak freely, then tidy the text by hand. Pair it with the Word Counter to check length, or the Grammar & Spell Checker for English proofreading. To hear text read back, use Text to Speech, which is the reverse of this tool.
Worked example: dictating a short note
With English (United States) selected and spoken punctuation turned on, you say: “send the report comma then call Priya question mark new line thanks period”. The tool replaces the punctuation words and capitalises the start of each sentence, giving “Send the report, then call Priya?” on the first line and “Thanks.” on the next.
A second phrase, “meeting notes colon budget approved full stop new paragraph next steps”, becomes “Meeting notes: budget approved.” followed by a blank line and “Next steps”. The counter above the box updates live, so you can see the word and character totals as you speak.
How it works and where the audio goes
The page uses the Web Speech API that browsers provide, called SpeechRecognition (or webkitSpeechRecognition in Chrome and Safari). The page asks for continuous recognition with interim results, so the browser returns guesses while you speak and a final result when you pause. Final results are added to the transcript; interim results appear in grey below it.
ToolWorld Online never receives your audio or text. The recognition itself is done by your browser’s vendor. In Chrome and Edge the audio is normally sent to Google’s or Microsoft’s online speech service, which is why an internet connection is needed and why accuracy and language support can change over time. Safari uses Apple’s speech recognition. Firefox does not offer the API to websites, so the tool shows an explanation instead of a Start button.
Common mistakes to avoid
Choosing the wrong regional variant lowers accuracy more than people expect. An Indian or Nigerian English speaker will often get better results with English (India) or English (Nigeria) than with English (United States), and the same applies to Spanish, French, Portuguese and Arabic variants.
Spoken punctuation is English-only and literal: if you say the word “period” meaning a time period, it becomes a full stop. Turn the option off when that matters. Finally, the transcript is not saved automatically. Copy or download it before closing the tab.
Names, brand names and technical terms are the words most often misheard, because the service guesses the most common word that sounds similar. Spell them out by editing the transcript afterwards rather than repeating them several times. If the wrong microphone is being used, for example a webcam across the room instead of your headset, change the default input in your operating system sound settings and start again.
Useful tips and limitations
- Speak in full phrases at a steady pace. Recognition improves when the service hears a whole sentence for context.
- Use a headset or sit close to the microphone in noisy places; background speech is the most common cause of errors.
- Pause briefly before saying a punctuation word; it helps the service treat it as a separate word rather than part of the sentence.
- Download the .txt file at natural break points during long dictation so a closed tab or browser update cannot lose your work.
Common questions
Does ToolWorld Online hear or store my speech?
No. ToolWorld never receives your audio or transcript. However, in Chrome and Edge the browser itself usually sends the audio to Google’s or Microsoft’s online speech service to recognise it, and Safari uses Apple’s speech recognition. Avoid dictating passwords or confidential details.
Why does speech to text not work in Firefox?
Firefox does not currently expose the Web Speech recognition API to web pages. Use Chrome, Edge or Safari, or type and paste text into the box instead.
Can I transcribe an audio file?
This tool listens to a live microphone only. To work from a recording, play it near the microphone, though results are usually less accurate than speaking directly.
Does it add punctuation automatically?
Some speech services add basic punctuation for some languages, and others do not. In English you can add it reliably by speaking punctuation words.
Why does dictation stop by itself?
Browsers end a recognition session after a period of silence or a time limit. While dictation is switched on the tool restarts it automatically; press Stop to end.
Is speech to text free?
Yes. There is no account, word limit or charge from ToolWorld Online.
ToolWorld Online does not receive or store your audio or transcript. Your browser’s speech service (for example Google in Chrome, Microsoft in Edge or Apple in Safari) processes the audio under that company’s privacy terms. The transcript stays in this tab until you copy or download it.