India Unveils Ambitious Plans for Affordable AI Model Powered by 18693 GPUs

India has unveiled its ambitious plans to develop a foundational AI model, aiming to rival established names like ChatGPT and DeepSeek R1. This initiative includes creating an affordable compute facility with 18,693 GPUs for startups and researchers to advance AI applications and algorithms. The move comes as DeepSeek's AI model recently surpassed ChatGPT on Apple's app store, challenging the dominance of US-based AI firms.

Representative image

IT Minister Ashwini Vaishnaw announced India's intention to launch a secure, indigenous AI model at a competitive cost. While global models charge between $2.5 and $3 per hour, India's model will be available for less than ₹100 (approximately $1.16) per hour after a 40% government subsidy. The Minister expressed confidence in developing a world-class foundational model that can compete globally.

AI Compute Infrastructure

The government has outlined the next steps in India's AI strategy, including offering 18,693 GPUs through empanelled bidders such as Jio Platforms, CMS Computers, Tata Communications, E2E Networks, and Yotta Data Services. These GPUs comprise 12,896 H100 units, 1,480 H200 units, and 7,200 MI 200/300 units. Additionally, eight projects have been approved under the AI safety institution initiative.

Addressing data security concerns related to DeepSeek, Vaishnaw stated that DeepSeek could be hosted on Indian servers after security checks. This would allow coders and developers to benefit from its open-source code. He also highlighted that India's foundational models would align with local contexts, languages, and cultures while eliminating biases.

AI Mission and Applications

Under the IndiaAI Mission, the government has allocated ₹10,372 crore to bolster the country's AI ecosystem. A key component is enabling 10,000 GPUs for AI Compute Infrastructure. The 18,693 GPUs will provide substantial computing power; for context, DeepSeek was trained on 2,000 GPUs and ChatGPT version 4 on 25,000 GPUs.

India aims to harness AI's potential to address large-scale challenges in healthcare, education, agriculture, logistics, and weather forecasting. The Minister announced that 18 applications have been selected for initial funding in areas like agriculture, learning disabilities, and climate change.

AI safety is a global concern, and the IndiaAI Mission prioritises safe and trusted models. Vaishnaw revealed plans for collaborative AI safety institutions operating on a hub-and-spoke model. These institutions will develop tools and frameworks necessary for AI safety.

AI Safety Initiatives

Approved projects for AI safety cover machine unlearning at IIT Jodhpur, synthetic data generation at IIT Roorkee, AI bias mitigation strategies, explainable AI frameworks at Defence Institute of Advanced Technology Pune, privacy-enhancing strategies at IIT Delhi and IIIT Delhi, an ethical certification framework for AI algorithms auditing tools, and an AI governance testing framework.

When questioned about potential impacts from US export restrictions on GPUs and AI chips, Vaishnaw emphasised India's reputation as a trusted nation globally. He noted that India respects intellectual property rights and has signed important technology agreements. This trust is expected to support the success of the IndiaAI Mission.

IndiaAI has also initiated a second round of Expressions of Interest (EoI) open to Indian academic institutes/organisations, autonomous bodies, R&D institutes/organisations, startups and companies. Themes include watermarking and labelling ethical AI frameworks risk assessment management stress testing tools and deep-fake detection tools.

India's focus on making modern technology accessible aligns with its economic vision. The Minister highlighted that six major developers or startups could build foundational models within eight to ten months or even sooner in four to six months. Algorithmic efficiency is crucial as it can deliver models at lower costs and faster than currently possible globally.

The government's commitment to enhancing India's AI capabilities reflects its determination to become a significant player in the global AI landscape while ensuring affordability and accessibility for all stakeholders involved.