here is maya1, our open source voice model:
We’re building the future of voice intelligence
@mayaresearch_ai team is incredible; amazing work by the team.
remarkable moment.
We’re launching a two week event for all IITs. Students at each IIT have the opportunity to unlock Perplexity Pro for their entire campus, and I’ll be visiting the IIT with the most Perplexity users later this year.
We're launching Veena TTS 🪕 on June 20
Our flagship text-to-speech model for Indian languages 🇮🇳
Natural, expressive, and actually sounds like us.
We’re launching two models:
Veena Lite
>Open-source and lightweight
>4 unique, natural-sounding voices
>The first open-source
here is maya1, our open source voice model:
We’re building the future of voice intelligence
@mayaresearch_ai team is incredible; amazing work by the team.
remarkable moment.
Introducing Indic Vault
The Largest Native Q&A Dataset for Indian Languages 🇮🇳
We're open-sourcing a high-quality, instruction-tuned dataset with Q&A pairs across major Indian languages.
Key highlights:
> Natural, conversational 2025-style language — not formal or outdated
>
Trained in deep learning with Yann LeCun.
Saw India’s AI bottleneck wasn't models — it was infra.
22, Dheemanth returned home to build Maya Research.
Open-sourcing Indic agents for the next billion.
Go build/support maya research - 🇮🇳/acc is here.
Hindi Speech Model should be free and open-source.
And we’re making it happen.
>High quality
>Multiple voices
>UI to generate audio + API for deployment
Follow me to get notified when it launches.
Listen to the audio till the end!
here is maya1, our open source voice model:
We’re building the future of voice intelligence
@mayaresearch_ai team is incredible; amazing work by the team.
remarkable moment.
We’ve got hundreds of DMs asking us why we chose to build Veena. Here’s our story:
Twitter is full of praise for American consumer voice apps like NotebookLM, Sesame, and Hume. @bharathkumar__1 and I strongly believe that the Indian consumer AI boom will be through a voice