In the frenzy of embodied intelligence with a daily investment of 300 million yuan, Innoangel has invested in half of the Tsinghua University alumni startups.
Track Frenzy: 3 Billion Yuan Flows in Daily
This wave of embodied intelligence fever is indeed scarily hot.
The "Report on the Development of Embodied Intelligence (2025)" jointly released by the China Academy of Information and Communications Technology and the Department of Electronic Engineering of Tsinghua University shows that in 2025, there were 744 investment events in the field of embodied intelligence and robotics in China, with a total financing amount of 73.543 billion yuan.
According to incomplete statistics from "Blue Shark Finance Society", as of the end of April 2026, there were more than 200 disclosed financing events in the field of embodied intelligence in China, with a total financing amount of over 35 billion yuan, which is equivalent to at least 300 million yuan being invested in the embodied intelligence track every day.
The magnitude of single - round financing and the frequency of large - scale financing are also continuously increasing.
According to incomplete statistics from "Blue Shark Finance Society", since 2026, there have been more than 20 single - round financings of 1 billion yuan or more in the embodied intelligence track. In the whole year of 2025, there were only 6 financings at this level, and in 2024, only Unitree Technology achieved this. In addition, many embodied intelligence companies have even received huge financings at a very early stage. For example, Lingchu Intelligence, an "embodied brain" company, raised 2 billion yuan in its Pre - A round; Tashizhihang completed a financing of 455 million US dollars in its Pre - A round, setting a new record for single - round financing in the domestic embodied intelligence field.
Investment and financing events with single - round financing of 1 billion yuan or more since 2026. Compiled by Blue Shark Finance Society
The financing rhythm is also accelerating rapidly. Since 2026, Zibian Robot has completed 3 rounds of financing, with a cumulative financing amount of over 3 billion yuan. Xinghaitu, Lingxinqiaoshou, Zhipingfang, Qianxun Intelligence, Xingdong Jiyuan, Jijia Shijie, Lingchu Intelligence and other companies have also completed 2 rounds of financing, and most of them have a financing amount of over 3 billion yuan. Looking at a longer time frame, Zhijian Power announced on March 9 that it completed 5 rounds of financing totaling 2 billion yuan within half a year of its establishment. Zhipingfang, Zibian Robot and other companies have completed a total of 7 rounds of financing since 2025.
Embodied intelligence companies that have completed multiple rounds of financing since 2026. Compiled by Blue Shark Finance Society
In the past many years, only a very small number of high - quality projects could complete three or four rounds of financing in a year. However, this rhythm has been compressed to three or four months in this wave of embodied intelligence fever. Wang Sheng, a partner of Inno Angel Fund, told "Blue Shark Finance Society": "I've never seen such a crazy state."
Frequent financing has promoted a rapid increase in valuations, and a batch of unicorn companies with a valuation of 10 billion yuan have been born intensively. Zibian Robot, Zhipingfang, Qianxun Intelligence, Xingdong Jiyuan, Xinghaitu and other companies have joined the ranks of unicorn companies with a valuation of 10 billion yuan after their latest round of financing. Wang Sheng revealed to "Blue Shark Finance Society" that among the embodied intelligence companies invested by Inno Angel Fund, the valuation of some projects has increased by 20 times in just a few months, and many embodied intelligence companies have brought dozens or even hundreds of times of book returns.
From "Missing Out" to "Hitting the Mark"
Among the investment institutions that have entered the market intensively, Inno Angel Fund has a strong presence. According to incomplete statistics from "Blue Shark Finance Society", since 2025, Inno has made nearly 30 investments in the field of embodied intelligence. It is one of the most active investment institutions in the track, and its early investment success rate is far ahead, being worthy of the title of "the No. 1 early - stage investment institution in embodied intelligence".
In the list of "Top 20 Most Investment - worthy Robot Companies in China" released at the 2025 World Robot Conference, 5 companies were first - round invested by Inno, including Jiasu Jinhua, Songyan Power, Yun Shenchu, and Qianjue Technology. In addition, Wang Sheng also led the investment of the alumni fund in Zibian Robot and Dongyi Technology.
Embodied intelligence companies in which Inno Angel Fund participated in early - stage investment. Compiled by Blue Shark Finance Society
However, this path is not easy. Instead, it is full of restrained and calm judgments, which has also led to many "misses" for Inno.
Wang Sheng told "Blue Shark Finance Society" that as early as the end of 2021, Inno Angel Fund had clearly defined two investment directions: large models and generative AI. It was one of the earliest angel investment institutions in China to focus on this field. At that time, there were few domestic companies really doing large models, and the Beijing Academy of Artificial Intelligence was the most representative, with core teams including Sun Maosong, Tang Jie, Liu Zhiyuan, and Wen Jirong.
Several leading large - model companies and several leading video - model companies had in - depth contacts with Inno during their first - round financing, but for various reasons, Inno did not invest in these projects.
In fact, until around 2024, the general judgment of large - model startup companies in the market was pessimistic. Most institutions believed that startup companies lacked revenue, suffered continuous losses, and would be squeezed out of the living space by large companies, and the money invested might be wasted. "At that time, we judged that large companies would definitely enter, and startup companies could not compete with large companies and would definitely be squeezed out," Wang Sheng said.
Looking back, these judgments were not accurate. For example, after Zhipu and MiniMax were listed on the Hong Kong Stock Exchange in January this year, their stock prices soared far beyond expectations: as of the close on May 18, Zhipu's stock price had reached HK$1,137, with a total market value of HK$506.9 billion - in just 4 months, the market value increased by nearly 10 times; MiniMax's stock price also reached HK$800, nearly 5 times the issue price of HK$165 per share.
Zhipu's stock price
Inno later reviewed and thought that it had not fully considered factors such as the macro - environment, new cycle, policy guidance, and premium of scarce targets, resulting in a deviation in the early judgment of the track. This means that even in a track that is optimistic in advance, cognitive blind spots are inevitable. "This may be related to the fact that Inno has always been engaged in early - stage investment," Wang Sheng said. "In this new era, many things have changed greatly, and we need to continuously learn and evolve."
Behind the High Investment Success Rate: The Key Word "Early"
So, why does Inno Angel Fund have such a high investment success rate in the embodied intelligence track? The answer can be summarized into two key words: early and in - depth.
Wang Sheng told "Blue Shark Finance Society" that as early as when formulating the investment direction plan for 2023 at the end of 2022, embodied intelligence had been clearly included in Inno's investment plan for 2024 - this was at least two years earlier than the market reached a consensus.
Regarding the investment logic, Wang Sheng gave three key judgments. First, the technological stage is clear. The large - model paradigm has been established, and embodied intelligence is in the innovation stage, rather than the pure scientific research stage. Second, the market space is large enough. The Internet and mobile Internet have mainly changed information circulation but have not significantly improved social productivity. Embodied intelligence returns to the physical world and can make up for this shortcoming, with a huge market scale. Third, startup companies have an advantage. Embodied intelligence involves long - chain links such as supply chain, manufacturing, training, and scenario implementation. Large companies are not willing to make heavy investments - they are better at and more willing to invest in short - chain problems that can be solved by throwing money and manpower, which leaves a huge space for startup companies. Generally speaking, the market of the AI Internet is 10 times that of the mobile Internet, and the market of the AI physical world will be 10 times that of the AI Internet. There will be many companies with a scale of hundreds of billions or trillions.
In 2023, Inno took the lead and invested in two robot body companies, Jiasu Jinhua and Songyan Power. These two companies have now joined the "Seven Little Dragons of Humanoid Robots".
Songyan Power became popular during the Spring Festival Gala with its 1:1 "Cai Ming Bionic Robot". Later, it quickly launched the consumer - grade high - performance humanoid robot "Xiaobumi" priced from 9,998 yuan, which became the best - selling product in the humanoid robot category on JD.com. Its founder, Jiang Zheyuan, plans to deliver 10,000 units in 2026.
The lightweight and portable small humanoid robot Booster K1 of Jiasu Jinhua has a cumulative shipment of over 1,000 units, and its monthly operating cash flow turned positive in December 2025, making it one of the few companies in the industry to enter the "commercial autonomy" stage. Thanks to its early entry at a low level, Inno has obtained dozens or even hundreds of times of book returns on these two projects.
Cai Ming's "same - style" bionic humanoid robot of Songyan Power (right). Screenshot from CCTV News
As of now, Inno Angel Fund has basically completed the layout of the entire industrial chain of embodied intelligence: from robot bodies (Jiasu Jinhua, Songyan Power), to the core algorithm layer such as Qianjue Technology (robot brain) and Manifold Space (world model), then to components such as joints and sensors, and the data layer such as Lingyu Intelligence (remote operation data) and Shouyi (hand operation data), covering the entire ecosystem. This way of layout of the entire industrial chain also forms a significant difference from the strategy of most VCs, which only invest in single star projects.
Deeply Rooted in the Academic Circle, Establishing the "Preferred" Status
The repeatedly verified experience for Inno to continuously invest in top - tier projects in the embodied intelligence track is that they established in - depth trust and long - term connections with university professors and doctoral students at a very early stage.
A notable phenomenon is that this wave of embodied intelligence entrepreneurship is significantly different from the mobile Internet and early AI tracks - the core entrepreneurs are no longer mainly former executives of Internet giants, and university scientific research teams have become the main force, especially professors and doctoral students from Tsinghua University have become the absolute main force. The core founding teams and chief scientists of the companies invested by Inno, such as Jiasu Jinhua, Songyan Power, Qianjue Technology, Lingyu Intelligence, Manifold Space, and Shouyi Technology, are all deeply associated with Tsinghua University.
Tsinghua - affiliated embodied intelligence companies. Compiled by Blue Shark Finance Society (incomplete)
Embodied intelligence belongs to bottom - level hard technology, and the core technology is in the hands of the academic and scientific research circles. When a professor or doctor with technology decides to start a business from the laboratory, they will prefer an investment institution that understands technology, has long - term communication, and mutual trust, rather than an institution that simply offers a higher price. This means that whoever can connect and take root in the academic circle earlier and more deeply can lock in higher - quality targets at an earlier stage.
Wang Sheng started to establish connections with these professors and doctoral students early on, not with the purpose of "I want to invest in you", but to learn, communicate, and explore together for a long time, and build trust with sincerity. "This is a snowballing process. At the beginning, you may only know a few teachers. As trust is passed on, you will gradually cover the top scientific research entrepreneurs in the industry and form a stable influence map."
This trust played a role at a critical moment. Wang Sheng told "Blue Shark Finance Society" that many embodied intelligence entrepreneurs actively approached Inno during their first - round financing, including Wang Qian (founder of Zibian Robot, undergraduate and postgraduate from Tsinghua University), Guo Yandong (founder of Zhipingfang, undergraduate and postgraduate from Beijing University of Posts and Telecommunications, doctor from Purdue University in the United States, former researcher at Microsoft headquarters, chief scientist at OPPO and XPeng), Han Fengtao (founder of Qianxun Intelligence, master in control theory and control engineering from Zhejiang University), Chen Jianyu (founder of Xingdong Jiyuan, doctoral supervisor and assistant professor at Tsinghua University), Gao Jiyang (founder of Xinghaitu, undergraduate from Tsinghua University), Hu Yuhang (founder of Shouxing Technology, doctor from Columbia University), etc.
In the list of "Top 20 Most Investment - worthy Robot Companies in China" released at the 2025 World Robot Conference, 13 companies had communicated with Inno Angel Fund during their first - round financing. This means that Inno's early access rate to top - tier projects in the field of embodied intelligence reaches 60% - 70%.
This "preferred" status among entrepreneurs has become the most solid barrier for Inno in early - stage investment in embodied intelligence. With the continuous expansion of its connection and influence in the track, a positive cycle has been formed in Inno Angel Fund: the earlier it seizes the resources of top - tier professors, the more the professor resources will actively recommend high - quality projects in return.
However, a high coverage rate does not mean investing in all projects. Some high - quality projects were missed for various reasons. To