The IndiaAI mission's empanelment process is proving to be a catalyst in reducing the costs of GPUs, essential hardware for artificial intelligence development. An analysis of the lowest (L1) prices from the second round of tendering for graphics processing units (GPUs) by the IndiaAI Mission reveals a price decrease of up to 10% compared to the first round. To ensure that the benefits of these lower prices are passed on to users, the first-round bidders will be asked to match these prices, resulting in substantial savings for users, according to the Economic Times.
The GPUs are being sourced under a government scheme aimed at providing subsidized computing power to local AI projects, making AI development more accessible to startups, researchers, and corporations. The decline in the price of the same GPU model from the first round to the second round ranged from just Rs 4 an hour, but the aggregate savings are expected to be significant.
Union Minister for Electronics and Information Technology (MeitY), Ashwini Vaishnaw, announced that seven companies will provide additional GPU capacity, bringing the total in the IndiaAI compute portal to 34,333 units. These units will be available at subsidized rates of ₹67 per hour, significantly below the average market rate of ₹115 per GPU hour. This pricing strategy makes India's GPU compute units notably competitive on the global stage, where rates can range from ₹213.44–₹256.12 ($2.5–3) per hour. The government is also offering a 40% discount on GPU access to startups, academic institutions, and students to further support innovation and learning.
The government has already empanelled 18,417 GPUs in the first tranche, with 14,000 units currently accessible through the IndiaAI compute portal, which became operational in March 2025. Vaishnaw also announced that a third round of GPU empanelment will be completed soon, indicating a continued expansion beyond the initial targets.
Rather than building government data centers, officials opted to empanel private providers who invest in the infrastructure while receiving 40% subsidies. This strategy has yielded bids already more than 40% below prevailing market rates before the additional subsidy was applied, according to a senior official.
The IndiaAI mission also includes the development of indigenous large language models (LLMs) optimized for India's 22 major languages and hundreds of dialects. Three startups have been chosen from 506 proposals to develop these LLMs, joining existing efforts such as Sarvam.
India is also approaching the 50,000 GPU export limit currently imposed by the United States. Efforts are underway to negotiate a relaxation of this cap as part of broader India–US trade discussions, which could significantly boost India's access to cutting-edge GPU technology.
In January 2025, India formally launched its ₹10,000-crore India AI Mission. Under this mission, empanelled bidders have offered 14,517 GPUs at L1 rates, against the target of 10,000 GPUs outlined in the IndiaAI compute pillar. It is expected that India will possess around 46,000 GPUs by June 10, upon the conclusion of the third round of bidding under the IndiaAI Mission. After the third round of bidding, MeitY will shift to a continuous empanelment process. Under this system, any company willing to match the lowest bid or per-hour cost for GPUs discovered during the first round of bidding will be eligible to supply these high-performance computing machines to participants in the ₹10,372 crore IndiaAI Mission.
