Google is inviting domestic AI-first startups between Seed to Series A that are solving uniquely Indian challenges or building specialised models, which can also be scaled up for global industrial applications.

Specifically, Google is focusing on startups building in four key areas, such as 1) Agentic AI for advanced automation workflows; 2) Multimodal AI, capable of processing audio, video, and image generation; 3) Physical AI, for smart manufacturing, robotics, etc and 4) Sovereign AI, which can develop and control its own artificial intelligence. This ensures independence in AI infrastructure, data, models, and talent.

Google startup accelerator programme will offer equity-free support, including collaboration with Google experts and access to advanced AI models, tools, including Gemini, Gemma, Imagen, Veo, Lyria, to solve complex technical roadblocks.

Further, startup professionals will also get access to free Cloud TPUs, the latest Google AI tools, and Google Cloud credits.

They will also get 1:1 guidance from specialists across Google DeepMind, Cloud, Health, Android, and more.

And, Google offers Weekly tracking with dedicated success managers to help startup founders complete the product and technology, and reach their growth targets.

The cohort is slated to kick off in late June with an intensive one-week in-person bootcamp in Bengaluru, concluding with a Demo Day in October.

Interested startup founders can submit their application on Google's official website (here) by April 19.

Several startups in the previous edition of the programme have seen significant progress in terms of product quality improvements, reach a wider audience globally and improve their revenue generation. For instance, Dview, which offers AI-powered analysis and enhanced cybersecurity to clients, reported 4x revenue growth. Whereas Superjoin utilised Gemini 3.0 to improve accuracy and latency by 50 per cent. And, Pulse identified $3M in at-risk revenue through large-scale feedback processing.