Google opens AI accelerator in India
Google launched its 2026 Startups Accelerator in India, inviting AI‑first startups for a three‑month, equity‑free program focused on agentic, multimodal, physical AI solutions and sovereign models. The program promises mentorship and resources — a clear runway if you're building agentic tools for real estate or local marketplaces. (blog.google) (prokerala.com)
Google published the announcement on March 31, 2026 and credited Farish CV as Regional Lead, AI Startup Programs — India. (blog.google) Each cohort will bring together roughly 10–15 startups in a hybrid mix of in‑person and remote sessions, and the program targets companies headquartered in India at the Seed to Series A stage. (startup.google.com) Google set an application deadline of April 19, 2026, and says the 2026 cohort will begin in late June with a week‑long in‑person bootcamp in Bengaluru and conclude with a Demo Day in October. (bestmediainfo.com) Selected teams are offered hands‑on access to Google models named in the announcement — Gemini, Gemma, Imagen, Veo and Lyria — and are eligible to apply for Google Cloud credits and free Cloud TPU access, subject to programme rules. (blog.google) Mentorship will be provided by specialists across Google DeepMind, Cloud, Health and Android, and participants receive structured OKR support including weekly tracking and dedicated success managers. (blog.google) Google cited concrete outcomes from prior cohorts: Dview reported 4x revenue growth, Superjoin used Gemini 3.0 to improve accuracy and latency by about 50%, and Pulse identified roughly $3 million in at‑risk revenue via large‑scale feedback processing. (blog.google) The announcement named alumni use cases: AiSteth screened more than 75,000 patients offline across 20+ states, Vaani AI processed over 1 million minutes monthly and cut manual calls by 75%, Resilience AI improved accuracy by 35% and reduced model run times by 23% with Gemma 12B, and VideoSDK cut costs ~40%, improved latency ~30% and doubled revenue after deploying a Gemini Live API voice agent. (blog.google)
