Star Map has raised an additional 2 billion yuan, with a valuation of 20 billion yuan

How does AI Gao Jiyang’s academic background help propel the company’s rise?

The investment community has learned that today (April 2), Xinghaitu announced the completion of a 2 billion yuan Series B+ financing round, and is still assembling a long roster of investors.

Industry giants such as Huadian Technology, Lens Technology, Silicon Investment, Shidai Bole, and AVIC Fund; long-term funds across primary and secondary markets such as Xiuyuan Capital, Hongzhang Investment, and Yuhai Capital; and state-owned institutions such as Financial Street Capital, Jinpu Investment, Beijing Kechuang, and Guoyuan Equity—along with PE funds including the funds under CICC Capital, PuHua Capital, Hongtai Fund, and Guangfa Qianhe.

By this point, Xinghaitu’s total financing has nearly reached 5 billion yuan. Its valuation has doubled in less than two months and has surpassed 20 billion yuan RMB.

“This year will be the first year we turn efficiency into results.” Xinghaitu CFO Luo Tianqi told the investment community. At the same time, the company is preparing to set up a fund, focusing on data technology and downstream application scenarios. It has already invested in nearly 10 companies.

Just now, Xinghaitu raised 2 billion yuan

Valuation doubled

Starting with the company’s founder, Gao Jiyang.

Born in 1992, Gao Jiyang earned his bachelor’s degree from Tsinghua University’s Department of Electronic Engineering. Later, he completed a Ph.D. in computer vision at the University of Southern California in only 3.5 years, setting a record for the shortest graduation time in his lab. After that, he worked at autonomous driving companies Waymo and Momenta.

In September 2023, Gao Jiyang led a young team with top academic backgrounds to launch a startup, throwing himself into the journey of embodied intelligence entrepreneurship. Xinghaitu was thus born, focusing on “one brain, multiple forms,” and it quickly sparked a fundraising frenzy.

Just two months after its founding, Xinghaitu completed the first delivery of angel-round financing. Afterwards, it raised funds fiercely, bringing in investors including IDG Capital, BV Baidu Venture Capital, Infinite Fund SEE FUND, Sequoia China Venture Capital, Seven Xi Investment, Hillhouse Capital, and Ant Group as the lead investor—along with investors such as MiHoYo, Wuxi Venture Capital Group, Tongge Venture Capital, Funplus, and others.

Particularly impressive is 2025: Xinghaitu completed five rounds of Series A financing in one go, with new backers including Kaihui Fund, Lenovo Venture Capital, Haier Capital, Today Capital, Meituan Longzhu, CICC Porsche Fund, Xianghe Capital, Beijing Robot Industry Development Investment Fund, and Yizhuang State-owned Investment, among other well-known institutions.

Entering 2026, Xinghaitu remains in full throttle. In February, it announced a 1 billion yuan Series B financing round. New investors include Jinding Capital, BAIC Industry Investment, Bihong Investment, Zhengxingu Capital, Qianhai Ark, and Yifeng Capital, making it yet another 10-billion-yuan embodied intelligence unicorn after Unitree, Zhiyuan, and Galaxy General—also on the list of top performers.

Until this time—less than two months later—it again raised 2 billion yuan. Xinghaitu’s valuation surpassed 20 billion yuan RMB, making it the embodied intelligence company with the highest valuation at present.

“We want to turn ‘spending’ from efficiency into results in 2026.” Luo Tianqi said. Over the past six months, Xinghaitu’s R&D expenses have been several times higher than since the company was founded. “We believe the industry’s Scaling (scalability) starting gun has already gone off.”

The results are also emerging. In 2025, Xinghaitu ranked first globally in market share, cumulatively serving more than 150 global well-known embodied intelligence developer partner companies, with top-tier coverage exceeding 90%. And this year, the company will build the world’s largest scale real-world embodied data set, continuously driving the evolution of embodied foundation models with million-hour real-world scenario data.

A milestone moment is that Xinghaitu has just released Fast-WAM, the fastest world model in the world. It proves for the first time: the power of world models is rooted in video modeling capability, not in the video generation process. This research completely bids farewell to the traditional inefficient world model paradigm of “imagine first, then execute.” By a disruptive reconstruction of the model’s underlying logic, it significantly improves world model inference speed, providing developers with an efficient path from prototype validation to large-scale deployment.

Turning Point

At this moment, embodied intelligence financing is boiling.

Since the beginning of the year, we have witnessed a funding storm in embodied intelligence. In the first quarter alone, the sector disclosed more than 40 financing events. The number of companies valued at 10 billion yuan+ has expanded to more than 10. Financing deals of more than 10 billion yuan occur frequently: in January, Variable Robotics completed a 10 billion yuan A++ round; in February, Zhujie Dynamics raised a 200 million US dollar B round, and ZhiPing raised a 10 billion yuan B round series; in March, Galaxy General secured a new round of 2.5 billion yuan financing, Jihua Vision obtained nearly 1 billion yuan in a Pre-B round… and everyone is quietly competing.

We are also about to witness the first wave of embodied intelligence listings. Not long ago, Unitree Technology’s IPO application for the Sci-Tech Innovation Board was accepted by the Shanghai Stock Exchange, with plans to raise 4.202 billion yuan. This means Wang Xing is finally taking Unitree to pursue the Sci-Tech Innovation Board’s “first embodied intelligence stock.” Meanwhile, Zhiyuan Robotics, Galaxy General, Zhongqing Robotics, Magic Atom, Leju Intelligence, Yunshenchu, and others have also updated their IPO progress. The teams are still growing, and 2026 may be a big year for embodied intelligence IPOs.

Capital is casting its votes with its feet. Behind this is a fairly common judgment: China’s embodied intelligence “ChatGPT moment” still needs five years, or even longer. That is precisely why, whether it’s financing in the primary market or IPOs, all parties are stockpiling ammunition for the long period of exploration ahead for the next several years.

And when capital, resources, and talent all converge on leading companies, the “Matthew Effect” intensifies. As this wave of financing heats up, eliminations are happening at the same time. According to third-party statistics, in the past year, more than 20 robotics companies worldwide have gone bankrupt, laid off staff, or divested their businesses—truly two worlds of ice and fire.

“Behind the prosperity of embodied intelligence, there are hidden concerns. Capital is accelerating its concentration toward leading companies and highly certain upstream core components.” Professor Wang Tianmiao of Beihang University said this earlier. Put in investors’ terms, “Out of more than 100 companies, only three or five remain in the end. Over 90% of the money and resources are sacrificed—there’s no way around it.”

As a result, the prospects for a closed commercial loop have become the watershed for how the industry positions itself. Only players with the ability to continuously self-generate cash flow can stay at the table for the long term.

It’s still that most straightforward principle: everything is driven by orders.