Key Highlights
BharatGen, an IIT Bombay-led government-backed AI initiative, is focusing on developing small language models (SLMs) alongside large language models (LLMs) to cater to domain-specific needs across sectors such as agriculture, finance, ayurveda, and legal
Recently awarded Rs 988.6 crore under the India AI Mission, BharatGen aims to create a sovereign AI ecosystem featuring multimodal LLMs with up to one trillion parameters and specialized SLMs that offer efficient and contextual solutions for Indian enterprises
The initiative is developing AI models for 22 Indian languages to enable inclusive, culturally relevant technology access, particularly benefiting rural and semi-urban populations
Strategy and Technical Approach
Smaller, distilled models derived from large foundational models will enable faster, cost-effective deployment in specific industries, unlocking enterprise value while supporting sovereign AI goals
BharatGen collaborates with industry bodies like Nasscom and global partners such as IBM to build scalable data pipelines, model governance frameworks, and benchmark standards tailored to India’s linguistic and commercial requirements
The project builds on BharatGen’s foundational Param-1 bilingual LLM launched earlier in 2025, trained on vast Indian language datasets to ensure high-quality outputs
Impact and Prospects
BharatGen’s inclusive AI ecosystem aims to democratize AI use across education, healthcare, agriculture, finance, and governance, fostering innovation with indigenous language and cultural context
This approach aligns with India's digital sovereignty push, ensuring self-reliance in AI while addressing the diverse linguistic landscape through efficient, smaller models specially fine-tuned for local needs
Sources: Economic Times, CIO Elets, IBM Newsroom