Voices

Every voice SangAI can speak with. They all run in your browser: a voice's model downloads the first time you play it, from this site or from Hugging Face, and stays in your browser's cache. Nothing you type here is sent anywhere.

Text Language for multilingual voices Speed Steps (Supertonic) Engine Status Filter

Voice	Language	Engine	Status

Licences

Piper voices: each voice carries its own licence on its model card; the engine is MIT.
Supertonic 3: model under OpenRAIL-M, browser client from Supertone under MIT.
Kokoro 82M: Apache-2.0.
Piper and Kokoro turn text into sounds with espeak-ng, licensed GPL-3.0.

Back to SangAI · Privacy