Maya-2, a voice model for Indian languages, is released with open weights

Maya-2, a voice (speech synthesis) model aimed at India, was published with open weights on October 27 under the Hugging Face name maya-research/maya2-native. The announcement on X says it covers more than 1.5 billion people across all 28 states and 8 union territories.

Claims from the announcement, not independently measured: it outperforms ElevenLabs v3, Sarvam Bulbul v3 and Cartesia Sonic 3.5. A reply from the same account says it builds on the Apache-2.0 Qwen3-1.7B-Base backbone and the Kyutai Mimi codec (CC-BY-4.0).

The post had about 230 likes by October 27. One account quoting it said the voice is really good but still has the familiar AI-voice feel.