India AI Mission Expands: Three More Startups Selected to Build Indigenous Foundation Models, Compute Capacity Doubled

In a significant step toward strengthening India’s AI capabilities, the government has selected three more startups — Soket AI, Gan AI, and Gnani AI — to develop indigenous foundation models under its India AI initiative. This move follows the earlier selection of Sarvam AI, marking a growing ecosystem of homegrown innovation in foundational AI technologies.

The announcement is part of India’s broader push to position itself as a global hub for artificial intelligence by fostering innovation, improving AI infrastructure, and promoting the development of models in local languages. The selected startups will focus on building large-scale AI models tailored to the Indian context, with a strong emphasis on inclusivity and regional diversity.

In a parallel boost to infrastructure, the government has also unveiled a major expansion of its AI compute capabilities. An additional 16,000 GPUs have been made available through industry partnerships, taking the total number of GPUs accessible to Indian startups and researchers to 34,000. This significant increase in computational power — made accessible via the cloud — will serve as a common AI platform for training and inference.

“This shared infrastructure is critical for enabling cutting-edge research and development in AI. It empowers startups and researchers across the country to innovate without being hindered by resource constraints,” said a senior official from the Ministry of Electronics and Information Technology.

The expanded compute facility is expected to accelerate the development of indigenous large language models (LLMs), speech models, and other foundational technologies that can address India-specific use cases, especially in sectors like agriculture, healthcare, education, and governance.

With this initiative, India reaffirms its commitment to creating a self-reliant AI ecosystem, ensuring that the benefits of artificial intelligence are accessible, inclusive, and aligned with national priorities.