Google DeepMind backs 16 climate-focused AI projects across Asia-Pacific

Google DeepMind has named the inaugural cohort of its Accelerator: AI for the Planet program in Asia-Pacific, selecting 16 startups, nonprofits, and research teams to receive mentorship and access to Google's environmental AI models. In its announcement, Google DeepMind said, "We've selected 16 organizations for our Google DeepMind Accelerator: AI for the Planet (APAC) cohort" to "help them advance their ideas using frontier AI models."

What's new

The 16-organization cohort is split across three focus areas. Six groups — 800 Trust, Kumi Analytics, Listening Lab, TelePIX, Wildlife.ai, and Yayasan Ekosistem Lestari — are working on protecting nature and building climate resilience. Five — Edufarmers, Living Roots, SIGMA, Terrastack, and X-Centric — are focused on sustainable agriculture. The remaining five — Archeda, City Syntax Lab, Climitra Carbon, Farmers for Forests, and Varaha Climate — are scaling climate and carbon-measurement solutions.

Participants get three months of mentorship plus access to Google's environmental model stack, including AnthroKrishi, ForestCast, and Perch, along with what Google describes as tailored technical support. The program opened with an in-person bootcamp in Singapore. Google's announcement does not disclose a funding amount attached to the program — the value on offer is model access, technical support, and mentorship rather than direct capital.

Context

This is a regional extension of Google DeepMind's broader "AI for the Planet" effort, which has previously run similar accelerator cohorts and paired them with purpose-built environmental models such as WeatherNext and ForestCast. Google has increasingly used these programs to put its research-grade climate and biodiversity models — many of which don't have an obvious commercial product path on their own — into the hands of smaller organizations that can apply them directly to conservation, agriculture, and carbon-accounting work in the field.

Why it matters

On its own, a 16-organization accelerator cohort is a modest, low-dollar move — Google didn't disclose a funding figure, and none of the 16 selected groups are household names. But it's a data point in a pattern worth tracking: the major labs are increasingly using free access to specialized models, rather than cash grants, as their preferred lever for demonstrating AI's non-commercial value. For Google specifically, pairing an APAC-focused cohort with home-grown models like AnthroKrishi and ForestCast doubles as a soft go-to-market motion — building real-world deployment case studies and regional goodwill for models that otherwise have thin adoption outside Google's own research and product teams.