New Delhi - [India AI Impact 2026 - February 18, 2026 – BharatGen, India’s flagship sovereign AI initiative, has announced the landmark launch of BharatGen Param2 17B MoE, a 17-billion-parameter multilingual Mixture-of-Experts (MoE) foundational modeI. This pivotal announcement, made at the high-profile India AI Impact Summit 2026, marks a defining step in the nation’s commitment to self-reliance and technological leadership in artificial intelligence.
The new model is the cornerstone of BharatGen's mission to build India-first, sovereign AI systems that fully reflect and capture the nation's immense linguistic, cultural, and societal diversity. Enabled through collaboration with NVIDIA, leveraging NVIDIAAI software and infrastructure, BharatGen’s end-to-end training pipeline is built on NVIDIA AI Enterprise, using NVIDIA NeMo open libraries such as NVIDIA NeMo-RL and Slurm workloads integrated with NVIDIA Base Command Manager to ensure performance and scalability.
Optimized specifically for multiple Indic languages, this initiative ensures that AI innovation is truly inclusive and accessible to every citizen and enterprise. Designed to serve critical use cases across core verticals, including governance, education, healthcare, agriculture, and enterprise solutions, the 17B Parameter Sovereign Multilingual AI Model is accelerating digital transformation across the country.
Speaking on the occasion, Rishi Bal, CEO, BharatGen said, “We are extremely delighted at unveiling PARAM 2 17B at the AI Impact Summit in Bharat Mandapam and also at the partnership with NVIDIA. As a truly sovereign AI initiative, the launch and unveiling of PARAM 2 17B marks a pivotal moment for India and we at BharatGen are elated at contributing to the country’s AI progress.”
In a move to foster a robust and open ecosystem, BharatGen is releasing its models, comprehensive documentation, and post-training workflows openly via its Hugging Face repository. This commitment empowers AI developers and enterprises to immediately begin building, fine-tuning, and deploying India-centric AI solutions that are built in India, for India, and ready for global impact.
BharatGen is a consortium under the Technology Innovation Hub at IIT Bombay and an Indian Government-Funded multimodal and large language model initiative for Indian languages, supported by the Department of Science and Technology (DST). Its mandate is to build AI for the nation by developing efficient AI models for Indian languages, creating a multilingual data repository, fostering public-private partnerships for scalable AI, and strengthening India's AI talent pool and startup ecosystem. BharatGen is building a family of LLMs and multi-modal FMs to serve diverse needs in India, including representing underserved languages, creating sovereign models for self-reliance, and partnering to build solutions for national importance across various sectors.