Text to Speech

About Text to Speech

This reads text aloud using the voices already built into your device and browser. It's useful for proofreading by ear (hearing a sentence often catches what the eye skips), following along with an article while doing something else, practising a script or pronunciation, or accessibility. Paste your text, pick a voice, and press play — with controls to pause, resume and stop.

How it works

It uses the Web Speech API's speech synthesis, which is part of your browser. The voice list is populated from the voices your operating system and browser provide, so what's available depends on your device. You can adjust:

• Voice — any installed voice, each tied to a language.
• Rate — how fast it speaks.
• Pitch — how high or low the voice sounds.

Playback runs on your device; you can pause and resume mid-sentence or stop entirely.

Assumptions and behaviour

• Voices come from your system/browser, not from us — the exact set (and their quality) varies between Windows, macOS, Android, iOS and different browsers.
• Rate and pitch are adjustable; the defaults are a normal speaking speed and pitch.
• Speaking a new text cancels any current playback, so you always hear the latest input.
• The reading is generated locally by the browser's synthesiser.

Limitations

• You can't download the audio. The Web Speech API plays speech; it doesn't hand back an audio file, so there's no MP3/WAV export — this is for listening, not producing an audio file.
• Voice availability and quality vary widely. Some systems ship only a couple of robotic voices; others have many natural ones. If a language's voice isn't installed on your device, it won't appear.
• Long text may be cut off by some browsers, which limit how much a single utterance will speak.
• It requires a browser that supports speech synthesis; if none is available, the tool tells you.
• Pronunciation of unusual names, abbreviations or code is approximate.

Privacy

Speech is generated by your browser's built-in synthesiser on your device — your text isn't uploaded by this tool. (Note that a few operating systems provide certain premium voices via a cloud service; the standard installed voices run locally.)