TLDR: The Indian government, under its IndiaAI Mission, is set to fund 15 to 20 indigenous AI models, including large and small language models, to foster local AI development. This initiative, announced at Semicon India, aims to localize core AI capabilities, provide affordable GPU access to developers, and deploy AI prototypes across various government sectors.
India is embarking on an ambitious journey to cultivate a robust homegrown artificial intelligence ecosystem, with plans to fund 15 to 20 indigenous AI models as part of its overarching IndiaAI Mission. This significant announcement was made by Abhishek Singh, Additional Secretary at the Ministry of Electronics and Information Technology (MeitY) and CEO of IndiaAI, during the Semicon India conference on September 2, 2025. The initiative underscores a strategic public-backed push to localize core AI capabilities, encompassing models, data, and computational resources, specifically tailored for Indian languages and contexts.
Already, four Indian foundation-model teams – Sarvam AI, Soket AI, Gnani.ai, and Gan.ai – have received government funding. Singh further revealed that ‘eight more, including LLMs and SLMs, are at the final stages of selection’ and are expected to be announced shortly. This aggressive approach aims to position AI as a critical productivity enhancer across diverse sectors such as agriculture, Micro, Small, and Medium Enterprises (MSMEs), manufacturing, and public services. Emphasis is also being placed on ensuring the safety, inclusivity, and sustainability of these AI advancements.
To provide substantial support to domestic AI developers, IndiaAI has facilitated access to ‘almost 38,000 GPUs’ for researchers and startups. These resources are available through tenders at an approximate cost of Rs 65 per GPU-hour (under $1), a stark contrast to global rates which typically range from $2.5 to $3. Singh highlighted that this initiative significantly alleviates a major constraint for local developers, enabling them to innovate and scale their projects more effectively.
However, the initiative also acknowledges potential challenges, particularly concerning energy and carbon footprints. Singh urged for dedicated green-power allocation and enhanced efficiency measures to ensure that the growth of AI does not compromise environmental sustainability.
In parallel with the funding for AI models, an update was provided on the deployment of government-funded AI prototypes. Approximately 30 such prototypes are currently being prepared for implementation across various ministries and states. The IndiaAI’s Application Development Initiative has seen 30 applications qualify in Phase 2, which are now undergoing technical hardening and departmental alignment. These applications are poised to be rolled out as state or ministry projects, addressing critical areas such as tuberculosis triage, crop advisory, disaster alerts, grievance redressal, and assistive learning.
Also Read:
- India’s AI Mission Secures Over 38,000 GPUs, Implements Voucher System for Compute Access
- India’s AI Kosh Initiative Crosses 1,000 Datasets, Bolstering National AI Infrastructure
Looking ahead, IndiaAI is organizing an ‘AI Impact Summit’ scheduled for February 2026 in Delhi. This summit will focus on key themes including ‘people, planet, progress,’ with dedicated working groups addressing jobs and skilling, economic productivity, social inclusion, democratizing AI resources (compute/data/models for the Global South), sustainable AI, AI-for-science, and safety. The summit will be preceded and accompanied by various events, including hackathons (with specific programs for women researchers), a startup pitch-fest, a research symposium, and an expo.
