India has taken a decisive step in its quest for technological sovereignty with the unveiling of BharatGen, a family of open AI models designed to operate across all 22 scheduled Indian languages, according to a report in India Today. 

The launch by IIT-Bombay at the Bharat Innovates 2026 event in Nice, France, comes at a moment when access to advanced AI systems is increasingly being shaped by geopolitical considerations. Reports of US authorities asking Anthropic to restrict foreign access to its most sophisticated models, Fable 5 and Mythos, have underscored the urgency for nations like India to build indigenous alternatives.

BharatGen is not a single chatbot but a suite of models tailored for diverse applications. At its foundation lies Param2, a text model capable of reasoning, coding and tool use, while functioning seamlessly across India’s linguistic spectrum. Complementing it are Shrutam2, which converts speech to text in multiple Indian languages, and Sooktam2, a text-to-speech system with zero-shot voice cloning capabilities.

Another model, Patram, is designed to process documents and forms commonly used in India, making it particularly valuable for banking, insurance and government services. Together, these models are being developed for use in governance, healthcare, education, finance, insurance and cultural preservation.

The timing of BharatGen’s debut is significant. With AI increasingly seen as a strategic technology, India is determined not to be dependent on foreign systems. The government’s IndiaAI Mission, a $1.2 billion initiative, has provided subsidised computing power to selected start-ups and research groups in exchange for publicly releasing their models.

BharatGen stands out as one of the flagship achievements of this mission, symbolising India’s ambition to build AI that reflects its linguistic and cultural diversity.

The project is led by Professor Ganesh Ramakrishnan at IIT-Bombay’s Department of Computer Science and Engineering, alongside CEO Rishi Bal and Vice President (Machine Learning) Dr Maneesh Singh.

More than 60 researchers, engineers and linguists from nine premier academic institutions are contributing to the effort, making BharatGen a truly national undertaking. This collaborative approach highlights India’s determination to build AI capacity rooted in its own institutions and talent base.

The unveiling also feeds into a broader debate about India’s place in the global AI race.

Infosys co-founder Kris Gopalakrishnan recently endorsed the view that India’s large IT services companies, including TCS, Infosys, Wipro and HCL Technologies, have not produced a ChatGPT-like competitor because their business models prioritise steady profits, job creation and foreign exchange earnings. 

Risk-heavy investments in cutting-edge AI research, which may take years to yield returns, are harder for such firms to justify. BharatGen, by contrast, represents a different model of innovation, one driven by academic institutions and supported by government funding.

India’s push to develop BharatGen reflects a wider ambition to reduce reliance on foreign AI systems and to assert its own technological identity.

With over 1.4 billion people and hundreds of languages and dialects, the country has unique needs that global AI models, often built around English and a handful of languages, cannot fully address. BharatGen’s multilingual capabilities are therefore not just a technical achievement but a strategic necessity, ensuring that India’s digital future is inclusive and self-reliant.

Agencies