Sakana AI Secures $135M to Build Japan’s AI Models

▼ Summary
– Sakana AI raised ¥20 billion ($135 million) in Series B funding, increasing its valuation to $2.65 billion from $2.5 billion pre-money.
– The startup, founded by former Google researchers, develops affordable generative AI models optimized for Japanese language and culture using small datasets.
– Funding will support R&D, AI model development, and expanding engineering, sales, and distribution teams in Japan.
– Sakana plans to expand beyond finance into industrial, manufacturing, government, defense, and intelligence sectors by 2026.
– The company focuses on creating sovereign AI solutions tailored to national cultures and values, differentiating from large-scale AI model competitors.
While global technology titans compete to build massive, all-purpose artificial intelligence systems, a new wave of specialized AI firms is emerging with a different focus. Companies like Sakana AI are developing models specifically designed for particular regions, industries, or unique applications, carving out a significant market niche. Tokyo-based Sakana AI has successfully closed a substantial ¥20 billion funding round, equivalent to approximately $135 million. This Series B investment values the company at a post-money valuation of $2.65 billion, a notable increase from its pre-money valuation of $2.5 billion.
Established in 2023 by former Google researchers Llion Jones, Ren Ito, and David Ha, Sakana AI specializes in creating affordable generative AI models that perform exceptionally well with smaller datasets. A core part of their strategy involves optimizing these models specifically for the Japanese language and its unique cultural context. The recent funding round attracted a powerful combination of investors, blending Japan’s financial leaders like Mitsubishi UFJ Financial Group (MUFG) with prominent international venture capital firms such as Khosla Ventures, Macquarie Capital, NEA, Lux Capital, and In-Q-Tel.
According to CEO David Ha, the new capital will be directed toward research and development for ongoing AI model creation. A significant portion will also fund the expansion of their workforce in Japan, particularly within engineering, sales, and distribution teams. This year has already seen the company strengthen its domestic position by forming partnerships with major Japanese enterprises, including Daiwa and MUFG, to develop tailored AI solutions. Ha, whose background includes leading research at Stability AI and roles at Google and Goldman Sachs, also revealed plans to broaden the company’s enterprise business beyond the financial sector. The goal is to move into industrial, manufacturing, and government sectors by 2026.
Looking to the future, Sakana AI has identified defense, intelligence, and manufacturing as key growth areas. The two-year-old startup also announced it will be “actively pursuing strategic investment, partnerships and M&A for long-term global growth.” The company perceives robust international demand for sovereign AI solutions that authentically represent national cultures and values. By concentrating its research and development on refining AI products after the initial large language model training phase, Sakana believes it can efficiently construct sophisticated models customized for Japan. This approach sets it on a distinct trajectory, separate from the race to build ever-larger foundational models.
This Series B financing arrives about one year after the company’s Series A round, which raised roughly ¥30 billion ($214 million) at a $1.5 billion valuation. Prior to this latest injection of capital, Sakana AI had raised a total of approximately $244 million as of March 2025. This new funding brings the company’s total funding to an estimated $379 million, providing substantial resources to execute its focused growth strategy.
(Source: TechCrunch)





