Google on Thursday unveiled a new Market Access Programme aimed at helping Indian startups bridge the gap between local pilots and global scale, alongside announcing additions to its Gemma open model family as part of its broader artificial intelligence push in the country.
“AI is moving out of research labs into classrooms, healthcare and hospitals, agriculture, factories, and across enterprises of all sizes,” Preeti Lobana, country manager for India at Google, said at the launch event. “AI startups are no longer experimenting at the edges. They are turning their capability into products that people use, trust and pay for. This is the point at which models turn into businesses.”
Bridging the Prototype Gap
The Market Access Programme is targeted at AI-first startups that have moved beyond the prototype stage and are preparing to scale responsibly. Google said the initiative would support founders on enterprise readiness through a specialised curriculum covering global enterprise sales, complex pricing structures and international buyer behaviour, while also offering facilitated access to a global network of chief information officers and senior executives.
The programme will also include global immersion modules in partnership with ecosystem organisations such as TiE Silicon Valley and Alteus, aimed at helping founders build in-person relationships in key buyer markets and international technology hubs. Applications for the programme are now open, the company said.
Expanding the Gemma Ecosystem
Alongside the market access initiative, Google announced new additions to its Gemma open model family, tuned to help startups build population-scale and production-ready AI applications. These include MedGemma 1.5, designed for advanced healthcare use cases that involve high-dimensional medical imaging at scale. The company said the model responds to growing demand for healthcare AI and enables developers to work with complex data such as CT scans, MRIs and pathology images.
Another addition, FunctionGemma, is a lightweight model optimised for function calling and supports on-device, agent-based systems. Google said the model allows AI applications to take secure and reliable actions locally, expanding the building blocks available to developers working on deployable, real-world solutions.
The announcements build on Google’s earlier commitment to invest $15 billion in India to develop AI infrastructure, including a Global AI Hub in Visakhapatnam, Andhra Pradesh. The facility is designed to provide a one-gigawatt foundation powered by green energy and Google’s advanced AI chips, giving startups access to high-performance compute resources.
According to the company, the latest launches align with its strategy of bringing physical AI infrastructure closer to developers in India, reducing latency and supporting faster scaling.