Sakana AI at a glance.
Read Oct 20, 2026. Valuations and amounts are copied as each source states them; a figure that is only reported (talks, or people familiar with the deal) is labeled as reported.
- Kind of company
- New frontier labsStart-ups building general AI models to rival the biggest labs.
- Based in
- Tokyo, Japan
- Founded
- 2023TechCrunch, Nov 17, 2025
- Last closed-round valuation
- $2.65B$2.65 billion post-money (TechCrunch, Nov 17, 2025)
- What it has shipped
- Has a product
- David Ha: former Google researcher; previously led research at Stability AI; also worked at Goldman Sachs (TechCrunch, Nov 17, 2025)
- Llion Jones: former Google researcher (TechCrunch, Nov 17, 2025)
- Ren Ito: co-founder (TechCrunch groups the three founders as former Google researchers) (TechCrunch, Nov 17, 2025)
Wealth-management AI with Daiwa Securities Group entering full-scale phase; separate integrated-analysis study for Japan's Ministry of Defense. Aug 2026
Sakana AI, Aug 2026
Shipped, or not yet.
“Has a product” is what our search found: something the public can use or buy, a model or API, research only, or nothing public.
The Darwin Gödel Machine (AI that rewrites its own code)
SourceShinkaEvolve, open-source framework for evolving algorithms with LLMs
SourceSakana Chat (with Fugu and Namazu models), Sakana Translate, Sakana Marlin research agent
SourceAI Lending Expert with MUFG Bank (being validated on live loan cases)
SourceFunding, as reported.
Every round we found, with the amount, valuation and investors as the source states them.
| When | Round | Amount | Valuation | Investors named | Source |
|---|---|---|---|---|---|
| Not stated | Series A | about ¥30 billion ($214 million) | $1.5 billion | None named | TechCrunch |
| Nov 17, 2025 | Series B | ¥20 billion (about $135 million) | $2.65 billion post-money | MUFG, Khosla Ventures, Macquarie Capital, NEA, Lux Capital, In-Q-Tel | TechCrunch |
What others have found.
- The AI Scientist paper was peer-reviewed and published in Nature (Sakana's page links nature.com/articles/s41586-026-10265-5); per Sakana, a fully AI-generated paper earlier passed peer review at an ICLR 2025 workshop and was withdrawn after acceptance. (Sakana AI, Mar 26, 2026)
Our note: Sakana's self-improvement research (Darwin Gödel Machine, AI Scientist) also fits the self-improving category. Sakana lists the AI Scientist's limits itself, e.g. 'It occasionally produces naive or underdeveloped ideas.' No 2026 funding round found in pages fetched.
