India is gearing up to introduce PARAM-2, a significant advancement in artificial intelligence, during the upcoming India AI Impact Summit 2026. Developed as part of the BharatGen initiative, PARAM-2 signifies India’s endeavor to create its own comprehensive AI models tailored to the nation’s languages, governance requirements, and cultural context.
Under the BharatGen program, which is backed by the Department of Science and Technology (DST), India has been steadily laying the groundwork to join a select group of countries capable of developing large-scale AI models. PARAM-2 is equipped to support all 22 Scheduled Indian languages and has been trained on datasets specific to India through Bharat Data Sagar.
Employing a Mixture-of-Experts architecture, PARAM-2 demonstrates efficiency in handling intricate multilingual tasks. Professor Ganesh Ramakrishnan from IIT Bombay, a key figure in the BharatGen initiative, emphasized that the launch of PARAM-2 signifies a collaborative effort among researchers, institutions, government entities, and industry stakeholders to empower India in shaping its AI trajectory.
In contrast to global consumer AI platforms like ChatGPT or Gemini, BharatGen operates on a distinct model. Rather than functioning as a centralized business-to-consumer service, its AI models are made available as national public digital assets. This approach enables various entities such as government departments, banks, hospitals, courts, and educational institutions to utilize them locally, even in secure offline environments, ensuring transparency, trust, and data sovereignty.
The foundation of BharatGen rests on the National Mission on Interdisciplinary Cyber-Physical Systems under DST, which initially allocated Rs 235 crore for the project. With further scaling supported by the IndiaAI Mission under the Ministry of Electronics and Information Technology, which has earmarked Rs 900 crore as part of a broader national AI strategy.
