All tools

Text to Speech

Read text aloud with your browser's voices

Have text read aloud using the voices built into your device, with speed, pitch and voice choice. The text stays in the page, but some voices are processed by your browser or OS vendor.

How to use Text to Speech

  1. Type or paste text (up to 30,000 characters).
  2. Pick a Voice. Voices are grouped by language, and each is marked “on this device” or “online”. Leave it on Browser default if unsure.
  3. Adjust Rate, Pitch and Volume, then press Play.
  4. Use Pause, Resume and Stop as needed. The word being spoken is highlighted when your browser reports word positions.

How it works

The tool uses the browser’s Web Speech API (speechSynthesis). Your text is split into sentence-sized chunks of at most 200 characters, and each chunk is queued as a SpeechSynthesisUtterance with the chosen voice, rate, pitch and volume. Splitting keeps long text working in browsers that stop long utterances after a while, and lets Stop and Pause react quickly.

The voice list comes from getVoices() and is refreshed when the browser fires voiceschanged. Word highlighting uses the utterance’s boundary events, which only some browser and voice combinations send. No speech is synthesised by this site: the voices belong to your browser or operating system.

Limits

  • Which voices exist, and how good they sound, depends on your browser and operating system. Some browsers list none until a voice is installed.
  • Word highlighting needs boundary events. Many online voices, and some browsers, do not send them, so no word is highlighted; speech still works.
  • Rate, pitch and volume changes apply from the next chunk, not mid-sentence. Pause and Resume are unreliable in a few browsers.
  • You cannot save the speech as an audio file: the Web Speech API does not expose the audio.
  • Text is limited to 30,000 characters.

Privacy

This site never sends your text anywhere and stores nothing. But speech is produced by your browser or operating system, and “online” (not local) voices may send the text to the vendor’s speech service, such as Google, Microsoft or Apple, to be turned into audio. Voices marked “on this device” run locally. If the text is sensitive, choose an on-device voice. Copy share link puts the text in the link’s # fragment, which browsers don’t send to servers, but anyone who has the link can read it.

Frequently asked questions

Why is the voice list empty?

Some browsers load voices a moment after the page opens, and some systems have none installed. The list updates by itself when voices appear. Installing a system voice or trying another browser usually helps.

What does “online” mean next to a voice?

The browser reports the voice as not local (localService is false), so audio may be generated on the vendor’s servers and the text sent there. This tool cannot see where it is processed.

Why is no word highlighted?

Highlighting relies on word boundary events that not every browser or voice provides. Try an on-device voice or another browser.

Can I download the audio?

No. The Web Speech API only plays sound and does not give access to the audio data.

More tools