Narrator: text to speech that runs in your browser

Narrator reads any text aloud with a speech model of your choice: Kokoro-82M’s 28 English voices (Apache-2.0), Supertonic 2’s ten voices in five languages (OpenRAIL), Meta’s MMS-TTS in ten languages (CC BY-NC 4.0), or any Kokoro, Supertonic or VITS model with ONNX weights you pick from Hugging Face. It runs on your GPU or, more slowly, the CPU, and nothing you type leaves the page.

How it works

  1. Type or paste up to 2,000 characters, or pick an example.
  2. Pick a model and a voice, and a language for models that speak several. Preview voice has it say hello.
  3. Press Load & read: the model downloads the first time, then reads aloud. Each reading is kept, to play again or save as a .wav.

Privacy and limits

  • Your text never leaves your device.
  • Each model downloads once and your browser keeps it.
  • Uses WebGPU where the browser offers it; elsewhere it runs on the CPU, more slowly.
  • Models from Hugging Face beyond the three built in are other people’s work and unchecked.

Models

Kokoro-82M · Supertonic 2 · MMS-TTS

Narrator runs on this site’s desktop, which needs JavaScript: with it on, this address opens the app. In the desktop’s terminal it’s app narrate. This page as Markdown.

Finder