Scylla's Band v4 is a new open-source text-to-speech model designed for speed and multilingual capabilities, capable of running on mobile devices. It features 14 distinct voices across six languages and offers advanced controls for intensity, pace, and pitch. Notably, the model can also sing by interpreting a JSON score, though this singing functionality is currently limited to English. AI
IMPACT Enables faster, more versatile TTS applications on edge devices and introduces singing capabilities to open-source models.
RANK_REASON Release of a new open-source model with novel capabilities. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →