Text to Speech

How to use text to speech

1. Type or paste your text in the input box.
2. Choose the Language (English, French or All languages), then pick a Voice. Voices on your device are listed first; voices marked “online” are listed after them.
3. Adjust the Speed, Pitch and Volume sliders if needed. The new values apply from the next sentence.
4. Click Read aloud. The text appears under Now reading with the current sentence highlighted and, when the voice reports it, the current word as well.
5. Use Pause and Resume to take a break, or Stop to end the reading.

Where the voices come from

This tool uses the Web Speech API built into your browser. It doesn’t bring its own voices: the list shows the voices your operating system and your browser provide, which is why it differs from one computer or phone to another.

• Windows: the Microsoft voices of the system (for example Microsoft David or Hortense). More languages can be added in Settings › Time & language › Speech.
• macOS and iOS: the system voices (Samantha, Thomas, Amélie…). Higher-quality versions can be downloaded in the Accessibility settings, under Spoken Content.
• Android: the voices of the system’s text-to-speech engine.
• Linux: usually the voices of speech-dispatcher, such as eSpeak.

If no voice exists for the chosen language, the browser’s default voice is used and a message tells you so.

Privacy: local and online voices

Most voices run entirely on your device. Some browsers, however, add online voices that send your text to the browser vendor’s servers to be synthesized. In Chrome on a computer, the voices whose names start with “Google” (such as Google US English or Google français) are online voices. In Microsoft Edge, the voices named “Online (Natural)” are online voices. The tool detects these through the browser and marks them with (online) in the list, and the note under the voice picker tells you which kind is selected.

For private or confidential text, choose a voice without the “online” label: it is synthesized locally and your text never leaves your device. The tool itself doesn’t send your text anywhere.

Long texts and highlighting

Chrome’s online voices tend to stop by themselves after about 15 seconds of continuous speech. To avoid this, the tool reads your text one sentence at a time, and splits very long sentences at a comma or a space so that each piece stays under about 180 characters. You won’t notice the switch, except for a very short pause between pieces.

The current sentence is always highlighted. Word-by-word highlighting depends on the voice: local voices usually report each word as it is spoken, while many online voices don’t, in which case only the sentence is marked.

Why there is no download

The Web Speech API plays speech directly through your speakers; it gives web pages no way to capture the audio. That is why this tool has no MP3 or WAV download. If you need an audio file, use your operating system’s own tools, such as a screen recorder with system audio.