India is stepping up its game in artificial intelligence, with the government pushing an ambitious plan to develop homegrown AI models within the next ten months.

The Modi administration has greenlit 18 proposals, offering support in the form of computing power, data, and funding to drive AI solutions in areas like agriculture and climate change. Technology Minister Ashwini Vaishnaw announced on Thursday that six major developers are on track to roll out foundational AI models by the end of the year.

“The AI models built in India will be world-class, capable of competing with the best out there,” Vaishnaw said, highlighting the country’s ambition to catching up with global AI leaders.

Boosting computing power

India is working towards building a computing capacity of over 18,000 graphics processing units (GPUs). Companies like homegrown E2E Networks Ltd. and Mukesh Ambani’s Jio Platforms Ltd. are in the race to develop this capacity, using advanced processing units like Nvidia’s H100 chips.

Vaishnaw also revealed that the average cost per AI computing unit is 111.85 rupees ($1.29) per hour. To help developers, the government will cover 40 per cent of the computing costs, making AI research and development more accessible.

IndiaAI mission and global partnerships

This initiative is part of the larger $1.2 billion IndiaAI mission, which aims to create both large and small language models. The push comes as China’s DeepSeek AI is making waves globally, competing with US giants like OpenAI.

India is also preparing to host DeepSeek on local servers, a move expected to enhance AI capabilities within the country.

Competing with global giants

Right now, the global AI scene is dominated by US tech heavyweights like Alphabet and Meta, but India is eager to carve out its own space in the industry.

With an estimated $30 billion in private sector investments set to flow into India’s data centres over the next few years, the country is laying the groundwork for a strong AI ecosystem.

(With inputs from Bloomberg)