Govt backs 20 indigenous AI foundation models; 237 projects get subsidised compute support
New Delhi, August 6
The government has identified 20 indigenous foundation model proposals for support under the IndiaAI Mission and approved 237 projects for subsidised AI compute support, as it continues to strengthen India's domestic artificial intelligence ecosystem and technological self-reliance.
In a reply in the Lok Sabha. Minister of State for Electronics and Information Technology Jitin Prasada, said the 20 selected proposals comprise 12 Large Multimodal Models and eight Small Language Models, with intellectual property rights to be retained by the applicants.
Among the outputs already launched are Sarvam AI's 30-billion and 105-billion parameter models and Indus, Gnani.AI's speech-to-speech model, BharatGen's multilingual foundation models and Avataar AI's video generation model.
The reply said 15 Compute Service Providers have been empanelled under the IndiaAI Mission, while 237 projects have been approved for subsidised compute support with 93.18 lakh GPU hours sanctioned. It also noted that a purchase order has been issued for a High-Performance AI Compute System of around 1.1 EFLOPS at the NIC Data Centre in Shastri Park, Delhi.
The government said the IndiaAI Mission seeks to build a robust and inclusive AI ecosystem while strengthening domestic capabilities in compute infrastructure, foundational models and advanced research.
According to the reply, 11 national-level hackathons and innovation challenges have been launched, resulting in 62 AI prototypes and 20 AI solutions being deployed across public sector institutions.
In addition, 58 AI Centres of Excellence have been approved across states and Union Territories, with 22 centres already approved and initiated across 13 states and UTs.
The reply also highlighted progress under the Semicon India Programme, stating that 12 semiconductor manufacturing projects, involving a total committed investment of Rs 1.64 lakh crore, have been approved so far, with three projects already commencing commercial production.
The government said these initiatives are aimed at building a self-reliant AI and semiconductor ecosystem, while reducing dependence on overseas technologies and strengthening India's long-term technological capabilities.
It also noted that, according to the Stanford Global AI Vibrancy Report published in November 2025, India ranks third globally in AI competitiveness and ecosystem vibrancy and is the second-largest contributor to GitHub AI projects.
— ANI
Reader Comments
Impressive numbers but I wonder about the practical implementation. 93 lakh GPU hours sounds great on paper, but how accessible will this actually be for smaller startups and academic institutions? Hope the distribution is equitable.
This is what Atmanirbhar Bharat should really be about! Not just manufacturing, but also building our own AI capabilities. Sarvam AI's models are already doing good work in Indian languages. Proud moment! 🙌
The semiconductor push is equally crucial. Approved 12 projects worth Rs 1.64 lakh crore - that's serious investment. Though I'd like to see more transparency on timelines since semiconductor fabs take years to become operational.
Great initiative but need to see the bigger picture. We are still behind in actual research output compared to US and China. Rank 3 in vibrancy is good, but let's not get complacent. Focus on foundational research, not just applications.
As someone working in this space, I appreciate the compute subsidy. GPU costs are prohibitive for most Indian researchers. This will democratize AI development. Good to see Gnani.AI's speech-to-speech model - localization is key for India!
58 AI Centres of Excellence across states is a good start, but
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.