The government has formally launched the India AI Mission with an initial outlay of ₹10,372 crore over five years, establishing a framework that combines public compute infrastructure investment, incentives for private AI model development, and a national AI applications programme targeting priority sectors.

At the centre of the mission is the IndiaAI Compute Platform: a distributed supercomputing infrastructure with an initial capacity of 18,693 GPU units hosted across six national AI Research, Development and Innovation centres at IIT Delhi, IIT Bombay, IISc Bangalore, IIT Madras, IIT Kharagpur, and a new facility in Pune.

The compute platform will be accessible to academic researchers, startups, and government departments on a priority basis, addressing what Indian AI entrepreneurs have repeatedly identified as the primary bottleneck to developing competitive large language models: the prohibitive cost of GPU access.

A dedicated ₹2,000 crore allocation targets the development of foundational AI models trained on Indian language corpora — covering the 22 scheduled languages plus approximately 150 additional languages with significant speaker populations. Current large language models developed outside India perform poorly on Indian language tasks, particularly in regional scripts.