Google DeepMind Is Launching an APAC Accelerator. The Market It Is Entering Just Collapsed by 44%.
Asia Pacific's climate technology sector is smaller than it was four years ago. Funding fell 44 percent between 2020 and 2024, according to KPMG and Google, even as the region attracted roughly $940 billion in energy transition investment in 2023 — about 45 percent of the global total. The money is flowing in. It is not reaching the startups trying to build with it.
Into that gap, Google DeepMind announced this week a three-month accelerator for Asia Pacific climate AI startups. The program offers mentorship and access to Google DeepMind's frontier models. It does not include direct investment. Selected teams attend an in-person bootcamp in Singapore and work with Google AI researchers on integrating models into real-world climate solutions, according to the announcement. Applications are open now. The company declined to say how many startups it plans to accept per cohort or what equity or intellectual property terms look like.
KPMG put numbers to what investors had been watching for two years. Agriculture, water, and waste — the sectors Asia Pacific's emissions profile depends on most — receive less than 10 percent of mitigation finance despite being among the region's hardest climate problems to solve. Sixty-five percent of climate tech deals in Asia are early-stage, according to the same report. Most startups at that stage lack the commercial infrastructure to apply frontier AI model access productively.
The mismatch is not abstract. A startup building AI-powered irrigation scheduling for Southeast Asian rice farmers may need three years and several million dollars in field trials before it has the engineering team and customer deployment to use frontier AI access. A three-month accelerator bootcamp does not solve that timeline. Growth capital does.
GreenTech companies take more than seven years on average to scale from Series A to meaningful commercial size, the KPMG report found. In a market where the real gap is exactly that kind of growth-stage capital, an accelerator offering model access instead of funding may serve Google's ecosystem interests more than the region's, according to climate tech investors and industry analysts who track Asia Pacific startup funding. They note that frontier AI model access is most valuable to companies with existing commercial infrastructure and engineering teams; early-stage climate startups in agriculture and water typically lack both.
Google DeepMind may have a separate calculation. Singapore is expected to announce new fintech and artificial intelligence policy this year, according to people with knowledge of the planning. Being embedded in the region's startup ecosystem before those announcements carries commercial value — a signal that Google's expansion into the region is not purely altruistic. A company spokesperson said the accelerator is not linked to any pending regulatory development.
What to watch next: whether the first cohort skews toward early-stage startups that cannot use frontier AI access effectively, or whether Google has quietly selected growth-stage companies with existing engineering teams. Google DeepMind more than doubled its APAC team over the prior year before launching this accelerator, a signal of how seriously it is treating the region. The bigger question — whether Google follows this program with actual check-writing — may matter more than the accelerator itself.