Supertone Releases Supertonic v3: On-Device Text-to-Speech Model with 31-Language Support, Fewer Reading Failures, and Expression Tags

Supertone has released Supertonic 3, its third ONNX-based text-to-speech model for on-device use.

Language support expands from 5 to 31, while repeat and skip errors are reduced and speaker similarity improves.

The model adds inline expression tags such as <laugh>, <breath>, and <sigh> for richer speech output.

It stays compact and CPU-friendly, making it suitable for local, browser, and edge deployment without a GPU.