Sakana AI secures $135M Series B for sustainable AI in Japan.

Sakana AI, a Tokyo-based AI R&D pioneer, has secured $135 million in a Series B round to scale its mission of building sustainable and sovereign AI for Japan.

The company got support from new and existing investors, including Mitsubishi UFJ Financial Group (MUFG), Khosla Ventures, Factorial, Macquarie Capital¹, Fundomo, Mouro Capital, New Enterprise Associates, Geodesic Capital, Lux Capital, Ora Global, MPower Partners, Shikoku Electric Power, and In-Q-Tel (IQT).

Around the world, funding for compute-intensive AI has surged, provoking discussions about profitability, energy use, and the path to practical deployments. Japan faces unique headwinds—an aging population, shrinking workforce, and finite resources—making a resource‑efficient AI ecosystem not just attractive but essential. Sakana AI frames its approach around operating with less, not chasing the hype of ever-larger models, aligning with a broader shift toward sustainable AI development.

Rather than training another giant model from scratch, Sakana AI has advanced techniques to blend open-source assets with proprietary systems. Its research agenda includes a self-improving architecture family, a multi-model collaboration framework, and innovative thinking beyond traditional transformers. The team has also explored autonomous AI scientists that can conduct research tasks end-to-end, aiming to reduce cycle times from hypothesis to publication while curbing compute demands.

In the market front, Sakana AI is turning its R&D gains into applied AI for Japan’s enterprises, starting with financial services and expanding into defense and manufacturing. The company notes a persistent gap between general models and the tacit knowledge embedded in professional workflows, which calls for deep, domain-specific engineering and post‑training customization to deliver tangible returns for customers.

The sovereign AI thesis envisions country-specific models that respect local culture, language, and regulation. Sakana AI argues that this post-training path can yield more relevant, safer, and cost-effective AI for national priorities than competing on upfront scale alone, positioning Japan as a laboratory for responsible AI deployment on a global stage.

Funds from the Series B will accelerate research into post‑training optimization, broaden applied AI collaborations, and build an ecosystem of partners and potential acquisitions to scale internationally from Tokyo. The push will also support talent recruitment across R&D, product, and corporate functions as the company scales its operations and partnerships with financial institutions and sector players alike.