Supertone has launched Supertonic 3, an updated on-device text-to-speech model that now supports 31 languages, a significant increase from its previous five. This new version boasts improved accuracy, reduced errors like repetitions and skips, and introduces expressive tags for inline control of emotions like laughter or sighs. Despite its expanded capabilities, the model remains compact at approximately 99 million parameters, making it suitable for deployment on resource-constrained devices without requiring a GPU. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
IMPACT Enhances on-device TTS capabilities, enabling more natural and expressive voice interfaces across a wider range of languages and devices.
RANK_REASON The cluster describes the release of a new version of an open-source text-to-speech model with new features and expanded language support. [lever_c_demoted from research: ic=1 ai=1.0]
