Google is taking a major step to advance India’s AI future with a large-scale expansion of its local AI hardware capacity, powered by its AI Hypercomputer architecture and the latest Trillium TPUs. The initiative aims to help Indian businesses, startups, and government organizations train and deploy advanced Gemini AI models within the country, reinforcing Google’s focus on AI sovereignty and data residency in India.
This means Indian developers and companies can now build and operate high-performance AI systems locally, reducing latency and supporting compliance with national data storage and privacy regulations.
Google stated that the goal is to make its most advanced AI tools available more efficiently to Indian users while ensuring complete control over their data. The company is introducing Gemini 2.5 Flash, its next-generation AI model, to India with local machine learning support, and providing early access to more powerful upcoming Gemini models — all with full data residency. This marks the first time Google Cloud is launching its top-tier AI models directly in India.
To better serve local users, Google is introducing new features tailored for India, including batch support for Gemini 2.5 Flash to enable cost-effective large-scale AI tasks, a preview of Document AI for automating document-intensive workflows, and a new capability called Grounding on Google Maps, which helps AI systems deliver precise, real-time location-based answers.
Beyond infrastructure, Google is investing in India’s AI talent through partnerships such as its collaboration with IIT Madras and AI4Bharat. Together, they are launching Indic Arena, an open platform that allows developers to test and benchmark AI models across multiple Indian languages. Google Cloud is supporting the project with cloud credits to promote inclusive AI research across India’s diverse linguistic ecosystem.
“Indic Arena will be a neutral benchmark for understanding how models perform across our many languages,” said Mitesh Khapra, Associate Professor at IIT Madras.
Google said it aims to empower startups, researchers, and enterprises to build AI “by Indians, for Indians,” inviting the country’s AI community, from universities to government bodies, to utilize its expanded Gemini capacity and sovereign-ready infrastructure.
