India to Launch Own Generative AI Model in 6-8 Months

India is preparing to create its own generative AI model, joining global leaders like OpenAI's ChatGPT and China's DeepSeek. The initiative, announced by Union IT Minister Ashwini Vaishnaw at the Utkarsh Odisha Conclave, will be supported by the India AI Compute Facility, which has acquired 18,693 GPUs to develop a Large Language Model (LLM) designed specifically for India.

Vaishnaw stated, "We believe that there are at least six major developers who can develop AI models in the six to eight months on the outer limit, and four to six months on a more optimistic estimate."

Vaishnaw emphasized that "a common compute facility is crucial for building a strong AI ecosystem," highlighting the need for researchers, startups, and academic institutions to have access to advanced computational infrastructure to drive AI progress. As part of the India AI Mission, the government has made the development of a shared computing resource a top priority.

Vaishnaw announced that the computing facility has surpassed initial expectations, securing nearly 19,000 GPUs, double the original target of 10,000. This includes 12,896 Nvidia H100 GPUs and 1,480 Nvidia H200 GPUs, among the most powerful AI chips on the market. Around 10,000 of these GPUs are ready for immediate use, with the rest being deployed gradually.

"This facility will be accessible to all and will be operational in the coming days," he confirmed.

Ashwini Vaishnaw stated that the average discount on AI Compute services is 42%, while high-precision compute units, essential for foundational models, are available at a 47% discount compared to market prices.

India is also introducing its own foundational AI model, aimed at addressing the country’s linguistic and cultural diversity while reducing biases in datasets. Vaishnaw announced that proposals for AI model development will be accepted starting today, with a projected development timeline of four to eight months.

"We believe that at least six major developers can build AI models within six to eight months, with some potentially delivering results in as early as four to six months," he said. The minister also spotlighted upcoming research in algorithmic efficiency, which could definitely lower the cost of model innovations.

A key priority of the India AI Mission is the implementation of AI applications to tackle significant societal challenges. Vaishnaw announced that 18 AI-driven projects have been selected for the first round of funding, focusing on three main areas: agriculture, climate change, and learning disabilities.

To promote responsible AI development, India will also set up an AI safety institution. Unlike other countries with centralized AI regulatory bodies, India will adopt a hub-and-spoke model, enabling multiple institutions to work together and offer safety tools and frameworks.