New Param2 17B MoE unveiled at India AI Impact Summit 2026
BharatGen India’s flagship sovereign AI initiative has launched its most advanced foundational model to date: the BharatGen Param2 17B MoE, a 17‑billion‑parameter multilingual Mixture‑of‑Experts model designed specifically for India’s diverse linguistic landscape. Revealed at the high-profile India AI Impact Summit 2026, the announcement marks a major advancement in the nation’s pursuit of self-reliant, culturally aligned artificial intelligence.
BharatGen has positioned Param2 17B as a cornerstone for India’s long-term AI strategy. Built in close collaboration with NVIDIA, the model leverages NVIDIA AI Enterprise, NVIDIA NeMo open libraries, and NVIDIA Base Command Manager for a fully optimized, end‑to‑end training pipeline. This ensures high performance, stability, and scalability capabilities essential for national‑scale AI adoption.
“PARAM 2 17B marks a pivotal moment for India, strengthening our path toward sovereign, inclusive AI.”
— Rishi Bal, CEO, BharatGen
What sets Param2 17B apart is its focus on Indic languages. With support across India’s linguistic spectrum, the model is engineered to power applications in governance, healthcare, agriculture, education, public services, and enterprise transformation. Its multilingual foundation ensures AI systems can understand and serve citizens across regions, bridging gaps often overlooked by global models.
BharatGen’s leadership emphasized the national significance of the launch. “We are extremely delighted at unveiling PARAM 2 17B at the AI Impact Summit in Bharat Mandapam,” said Rishi Bal, CEO of BharatGen. “As a truly sovereign AI initiative, the launch marks a pivotal moment for India, and we are proud to contribute to the country’s AI progress.”
In line with India’s commitment to open and collaborative innovation, BharatGen is releasing its model, documentation, and post‑training workflows openly via its Hugging Face repository. This move empowers researchers, startups, and enterprises to fine‑tune and deploy India-first AI systems immediately expanding accessibility and accelerating innovation across sectors.
Supported by the Department of Science and Technology (DST), BharatGen continues to build a family of language and multimodal models that reflect India’s linguistic diversity and drive national AI capability. With Param2 17B, India takes a decisive step toward a sovereign, globally competitive AI future.