{
  "id": 19279964,
  "title": "Sakana AI Targets Record Japanese IPO, Seeking to Top Mercari's 2018 Benchmark",
  "url": "https://urgent.news/2026/11/10/sakana-ai-targets-record-japanese-ipo-seeking-to-top-mercaris-2018",
  "topic": "ai",
  "section": "AI",
  "published": "2026-11-10T04:16:51.000Z",
  "source": {
    "name": "BusinessKorea",
    "slug": "businesskorea",
    "url": "https://www.businesskorea.co.kr/news/articleView.html?idxno=280896"
  },
  "original_language": "en",
  "account": "Sakana AI, a Japanese generative AI startup, is gearing up for a potential IPO as early as fall 2027, marking a potential record for Japanese startup listings. If successful, its first-day market capitalization could top the 680 billion yen ($4.3 billion) benchmark set by rival Mercari in 2018. The IPO's specifics, including venue, timing, offering size, and valuation, are still under discussion.\n\nSakana AI is partnering with U.S. investment bank Goldman Sachs as its IPO adviser and is in talks with Japanese brokerage firms Daiwa Securities and SMBC Nikko Securities. The company is considering listings on both the Tokyo Stock Exchange and Nasdaq in the United States, with a projected IPO date of fall 2027.\n\nFounded in 2023 by former Google AI researchers David Ha and Llion Jones, and Mercari executive Ren Ito, Sakana AI quickly attained unicorn status, achieving a valuation of over $1 billion just a year after its inception. A funding round led by Mitsubishi UFJ Financial Group (MUFG) valued the company at approximately 400 billion yen in November 2025.\n\nThe startup's primary goal is to develop Japanese AI models and expand its sales. It is currently working on its own model, Sakana Sawara, which it plans to release in spring 2027 after launching a smaller version. Sakana AI aims to generate annualized revenue exceeding 30 billion yen in 2027, roughly ten times its projected 2026 revenue.\n\nThe company's strategy focuses on achieving AI sovereignty by controlling its technology and operations within Japan. Unlike U.S. tech giants, Sakana AI combines existing open-source models and applies evolutionary techniques to perform well with fewer computing resources, targeting business-to-business clients such as financial institutions, corporations, and government agencies. Its AI product, Sakana Fugu, which combines existing models, has already been made available to the public, but its capabilities are still dependent on the performance of foreign-developed foundation models.\n\nTo overcome this limitation, Sakana AI is developing Sakana Sawara, which aims to surpass existing models by independently pre-training on large-scale datasets. The company has already secured major Japanese clients, including MUFG Bank, Sumitomo Mitsui Banking Corp., and Sumitomo Corp., and is seeking to expand its reach in defense and cybersecurity, where reduced reliance on foreign technology is a growing concern.\n\nHowever, global investor sentiment remains a potential risk for Sakana AI's IPO, with uncertainty surrounding AI regulation, data center investment, and the strategies of other major AI companies, such as Anthropic, following the U.S. midterm elections. The company's ability to demonstrate the performance and commercial viability of its own model, particularly its independent pre-training capabilities, will be crucial in securing investor confidence for its IPO.",
  "summary": "Japan's generative AI startup Sakana AI is preparing for a potential listing as early as fall 2027, in what could become one of the largest IPOs in the country's startup history. If completed, it would be the first listing by a Japanese company specializing in generative AI. Its market capitalizatio",
  "key_points": [],
  "editors_take": "Sakana AI’s pursuit of a record-breaking listing signals a shift toward domestic AI sovereignty as the startup attempts to decouple Japanese enterprise technology from a reliance on foreign-developed foundation models.",
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}