Feed
Maya Research Launches Open-Weight TTS Model for 11 Indic Languages
AI

Maya Research Launches Open-Weight TTS Model for 11 Indic Languages

27 Oct 2026, 6:03 pm · 2h ago · 1 min read · Analytics India MagazineNew

Maya Research has released Maya 2 Native, an open-weight text-to-speech model supporting English and 11 Indic languages, including Hindi, Tamil, Telugu, Bengali, Marathi, and Kannada. Distributed under an Apache 2.0 open-source license, the model delivers 24 kHz mono audio streamed in 80-millisecond frames for low-latency voice assistants and regional AI applications. Maya 2 Native features 20 built-in synthetic voices and offers cross-lingual voice cloning using a 3-to-10-second reference audio sample, alongside ready-to-deploy inference code and an OpenAI-compatible speech endpoint.