In 2026, capital is "going all out" and has nurtured two more robot unicorns.
The festive atmosphere of the Spring Festival has not yet faded, and several good news have come from the embodied intelligence track. On February 24, Qianxun Intelligence, a leading enterprise in embodied intelligence, announced that it had completed two rounds of financing totaling nearly 2 billion yuan recently, setting a new high in the field of embodied intelligence since 2026.
Image source: Screenshot from the official WeChat account of "Qianxun Intelligence"
Qianxun Intelligence is not an isolated case. Since 2026, the financing fever in the field of embodied intelligence has been continuously rising. Not only have many enterprises received huge amounts of financing, but some are also preparing for IPO. As capital accelerates its layout, the opportunities and challenges in the embodied intelligence track have become increasingly clear.
1. Robots "Show Off Skills" on the Stage, and Capital Accelerates Layout
Behind Qianxun Intelligence's 2-billion-yuan financing, the list of investors is quite impressive. There are not only super-first-tier institutions such as Yunfeng Capital and Hundi Investment (Ge Weidong), but also industrial capitals such as Synstellation Capital, TCL Venture Capital, and Minghui Zhiyuan, state-owned capitals such as Chongqing Industrial Investment Mother Fund and Hangzhou Jintou, as well as strategic investment institutions such as 360 Fund and Houxue Capital. Meanwhile, old shareholders such as Shunwei Capital, Prosperity7, Fortune Capital, and Bairui Capital are also continuously increasing their investment.
After this round of financing, the valuation of Qianxun Intelligence will exceed 10 billion yuan.
One day earlier, on February 23, Zhipingfang, the global leader in robot basic models, announced the completion of its Series B financing, with a scale of over 1 billion yuan. Zhipingfang stated that after this round of financing, the company's valuation exceeded 10 billion yuan, firmly ranking it in the first echelon of domestic embodied intelligence enterprises. The investors in this round cover various forces including Internet and AI giants, leading central state-owned enterprises, and the leading enterprise in the Tesla ecosystem. The investors include Baidu, CRRC, Yuxin Technology, Senqilin, Yunbai Capital, Guotai Haitong, etc.
It is worth mentioning that in the past year, Zhipingfang has completed a total of 12 rounds of financing, making it the fastest - financing embodied intelligence enterprise globally.
Image source: Screenshot from the official WeChat account of "Zhipingfang Technology"
In addition, on February 11, Xinghaitu completed a nearly 1-billion-yuan Series B financing. This round of financing gathered the support of industrial capitals, international funds, and old shareholders such as Golden Ding Capital, BAIC Capital, Bihong Investment, Zhengxingu Capital, Qianhai Ark, Yifeng Capital, Cathay Capital, Meituan Longzhu, Today Capital, Xianghe Capital, and Hillhouse Venture Capital.
On February 10, Daxiaorobot, a subsidiary of SenseTime, announced the completion of its angel - round financing. This round of financing was led by Ant Group, followed by institutions such as Qiming Venture Partners, Jinjing Capital, Hongyi Investment, Lenovo Capital, and Hanyuan Assets of Shanghai Jiao Tong University Mother Fund. The old shareholder, Shangtang Guoxiang Capital, continued to increase its investment.
According to incomplete statistics, there were more than 20 financing events in the global embodied intelligence and humanoid robot fields in January this year, with the total disclosed financing amount exceeding 16 billion yuan.
In terms of IPO, Estun, a leading enterprise in industrial robots that was listed on the main board of the Shenzhen Stock Exchange in March 2015 and submitted its prospectus to the Hong Kong Stock Exchange again in January this year, has recently passed the hearing of the Hong Kong Stock Exchange and plans to be listed on the main board of the Hong Kong Stock Exchange.
Enterprises such as Unitree Robotics, Leju Robot, and DeepClouds have disclosed their IPO guidance progress and plan to enter the capital market in 2026.
In January this year, Gu Shitao, the co - founder of Magic Atom, once revealed that "there may be new news in the secondary market as early as 2026, and the company is arranging the listing schedule at the fastest speed."
The intensive financing and IPO preparations have provided financial support for the embodied intelligence industry. Judging from this year's Spring Festival Gala stage, these investments are being transformed into visible results.
On the stage, domestic robots of multiple brands such as Unitree Robotics, Songyan Power, and Magic Atom appeared. There was a humanoid robot that highly resembled Cai Ming, hundreds of quadruped robots MagicDog performing in panda costumes, and the MagicBot Z1 dancing on stage with Yi Yangqianxi, Jordan Chan, Luo Jiahao, and Jerry Yan...
Image source: Video screenshot
This year's robots showed a higher level of flexibility and intelligence than last year, which is the result of the combined effect of capital investment and industry technology accumulation. Now, a clear chain has been formed in the embodied intelligence industry: capital injection supports R & D breakthroughs, product progress enhances market confidence, which in turn attracts a new round of capital attention. As this cycle continues to operate, the technological breakthroughs and commercial implementation of the robot industry are expected to accelerate simultaneously.
2. Cold Thinking under the Up - Surge, Robots Still Need to Overcome Challenges
It is worth noting that although there have been technological breakthroughs and the financing fever continues to rise, the embodied intelligence industry is still in its "infancy." Data from the GGII shows that the global market scale of embodied intelligent humanoid robots was 6.339 billion yuan in 2025, and it is expected to exceed 100 billion yuan around 2030.
From "infancy" to "adulthood," the embodied intelligence industry still faces challenges in terms of technology, cost, and regulation.
Domestic first - tier R & D experts once put forward four technical pain points for the mass production of humanoid robots at a series of teleconferences on humanoid robots:
1. The integration of hand and leg hardware is difficult, requiring higher peak loss power and driving ability;
2. It is difficult to ensure the stability of the leg walking algorithm, and there is no real human - like walking algorithm in the academic circle;
3. The fine perception of the hand is difficult, involving 21 - 26 degrees of freedom of the hand;
4. Algorithms such as dual - arm coordination, hybrid intelligent control, and avoidance are difficult and require high - dimensional planning.
Image source: Canstock Photo
Take the humanoid robot resembling Cai Ming at this year's Spring Festival Gala as an example. Although its appearance is quite realistic, it still seems rather mechanical when interacting with people and cannot conduct natural and smooth communication. This shows that there are still many technical thresholds to cross from "looking like a human" to "being able to interact like a human."
In fact, this is also the current situation of many robot products - more for performance than for practical applications. As the industry gradually matures, enterprises need to shift more energy from "showing off skills" to "practical use" and make robots truly useful in factories and in daily life.
Secondly, in terms of cost. Nowadays, the prices of most robots are not low. In the "NOETIX JD.com flagship store," the price range of the Xiaobumi, the same model as the one at the Spring Festival Gala, is from 9,998 yuan to 21,600 yuan; the price of the humanoid robot, the same model as the one at the Spring Festival Gala, is 498,000 yuan. In the "Unitree JD.com flagship store," the take - home price of the G1 EDU U2 advanced version humanoid robot is about 208,600 yuan.
Image source: Screenshot from JD.com
The mass production of humanoid robots also needs to consider the cost control of software and hardware and how to gradually average the cost with the scale effect.
Many materials and components of humanoid robots require customized design solutions. The high - precision motors required by intelligent agents are different in structure and materials from commonly used motors, and standard components are difficult to integrate directly, making it even more difficult to reduce costs.
Hidden behind the high cost and high technology, the real difficulty in the mass production of humanoid robots is to make a choice between technological leap and mass - production time.
If you want to produce a very "human - like" intelligent agent, it will take more time. If you mass - produce a less "human - like" one first, the advantage of differentiated competition with the existing intelligent agents in the market will not be so great. However, there is also a time - difference competition. The fear is that while some enterprises are patiently conducting R & D, their competitors have taken the lead in mass - producing a low - end version to capture the market mindshare.
Therefore, it can also be said that the real difficulty in the mass production of humanoid robots is "time."
As robots enter scenarios such as factories and families, regulatory issues have also become an urgent topic to be addressed. For example, how to ensure data security, how to protect user privacy, and where the ethical boundaries of robots lie.
The current development of the embodied intelligence industry is still in its initial stage, like a child who has just learned to walk. The movements are not yet stable, and there are not many things it can do, but the prospects are promising. From its budding to its appearance on the Spring Festival Gala stage and then to the capital up - surge, the Chinese embodied intelligence industry is moving forward steadily. Given time, this industry is expected to yield more fruitful results.
This article is from the WeChat official account "Fun Interpretation of Business" and is published by 36Kr with authorization.