Humanoid robots are competing to become unicorns.
Before and after the Spring Festival, the competition in the embodied intelligence track was fierce. Besides competing on the Spring Festival Gala stage, projects off - stage were also in a race for financing.
After the Spring Festival holiday, three companies in the embodied intelligence track - Zi Bianliang, Qianxun Intelligence, and Zhipingfang - almost simultaneously announced the completion of their latest financing rounds and were promoted to unicorns with a valuation of over 10 billion yuan. Just before the holiday, Xinghaitu also released financing news. After completing a Series B financing of 1 billion yuan, its valuation reached 10 billion yuan. That is to say, within a holiday period, four new unicorns in the embodied intelligence industry emerged in the public view.
It is worth mentioning that all four companies belong to the "embodied brain" direction in the track.
Financing Race in the Embodied Brain Field
These four newly - emerged embodied unicorns all belong to the brain direction, but they differ in terms of technical routes, product forms, and commercialization focuses.
First, let's look at Zi Bianliang. After announcing in January that it had received over 1 billion yuan in financing from institutions such as ByteDance and Sequoia Capital China, the company announced the completion of a new round of financing of hundreds of millions of yuan. This round was led by industrial capitals such as SAIC Financial Holdings and CICC SAIC Fund, with old shareholders such as Meituan Longzhu, Sequoia Capital China, Beijing Information Industry Development Investment Fund, and China Development Financial Services increasing their investments.
It is reported that after this round of investment, Zi Bianliang became the company with the highest valuation among the embodied intelligence companies that have announced financing in 2026. According to public information, since its establishment more than two years ago, Zi Bianliang has completed about 10 rounds of financing, with a total financing amount of over 3 billion yuan.
Wang Qian, the founder of Zi Bianliang, studied in the Department of Electronic Engineering at Tsinghua University for his undergraduate degree. He began researching in the field of AI in 2009 and then switched to the field of robotics during his doctoral studies. In terms of technical direction and business progress, Zi Bianliang has currently launched a native multi - modal input - output architecture called WALL - A; at the same time, it has launched the Quantum One and Quantum Two wheeled dual - arm robots, achieving self - research on core components such as joints, drives, and controllers.
Next, let's look at Qianxun Intelligence. Recently, it officially announced the completion of two consecutive rounds of financing totaling nearly 2 billion yuan, with a post - investment valuation of over 10 billion yuan. The investment institutions included Yunfeng Fund, Hundun Investment, Sequoia Capital China, Chongqing Industrial Investment Mother Fund, Hangzhou Jintou, Shunwei Capital, Qiancheng Capital, and Prosperity7, among other different types of investment platforms.
According to Qichacha information, as of now, Qianxun Intelligence, founded in January 2024, has completed 6 rounds of financing. In addition to the above - mentioned institutions, leading investors such as Huafa Group, Hubble Investment, and Orient Fortune Capital also appear in Qianxun Intelligence's shareholder list.
Han Fengtao, the founder of Qianxun Intelligence, has a background from the School of Control at Zhejiang University. He co - founded the collaborative robot company Luoshi and served as its CTO; co - founder Gao Yang is currently an assistant professor at the Institute for Interdisciplinary Information Sciences at Tsinghua University. From public information, the main line of Qianxun Intelligence is brain/VLA + data scaling + adaptation to the robot body.
In June 2025, the company released the Moz1 humanoid robot; at the end of the same year, the company said that the humanoid robot was put into operation on the embodied intelligence production line at CATL's Zhongzhou Base and entered the scenario of JD Retail stores. In terms of model capabilities, in January this year, the company open - sourced the Spirit v1.5 model and topped the RoboChallenge list.
The founding teams of Zhipingfang and Xinghaitu both have backgrounds in autonomous driving.
Zhipingfang was founded by Guo Yandong in Shenzhen in early 2023. Before that, Guo Yandong had held core R & D positions at Microsoft, XPeng Motors, and OPPO. During his tenure at XPeng Motors, he was responsible for the R & D of intelligent systems based on deep learning as the chief scientist, promoting the mass - production and implementation of deep - learning technology in automotive - form hardware terminals.
In the latest Series B financing round, Zhipingfang raised over 1 billion yuan, and its valuation also exceeded 10 billion yuan. The investors included CRRC Capital, Guotai Haitong, Baidu, and Chengdu Science City Venture Capital. Qichacha shows that as of now, Zhipingfang has completed 5 rounds of financing, and there are also a series of shareholders such as Shenzhen Capital Group, Nanshan Strategic Emerging Industry Investment, Cornerstone Capital, Bloomage Biotechnology, and Guozhong Capital.
Zhipingfang is one of the earliest domestic embodied intelligence companies to explore the end - to - end VLA technology direction. It has currently launched its self - developed all - domain whole - body VLA large model GOVLA, emphasizing the ability to migrate across multiple scenarios while achieving unified control of the robot's whole - body movements. In terms of commercial implementation, in September 2025, Zhipingfang reached a cooperation with the Huike system, planning to deploy over 1,000 robots in three years, with an order amount close to 500 million yuan.
Gao Jiyang, the founder and CEO of Xinghaitu, worked at Waymo and Momenta after graduating with a doctorate, being responsible for the R & D and mass - production implementation of core intelligent driving systems such as visual perception, positioning, and control; several members of the core team also come from leading autonomous driving companies.
The latest 1 - billion - yuan Series B financing round of Xinghaitu also attracted a series of institutions such as Jinding Capital, BAIC Capital, Bihong Investment, Zhengxingu Capital, Qianhai Ark, and Meituan Longzhu. Qichacha shows that since its establishment in September 2023, Xinghaitu has completed a total of 8 rounds of financing, and other shareholders include Ant Group, IDG Capital, Beijing Government - guided Fund, Hillhouse Ventures, and Mihoyo.
In terms of technical progress, Xinghaitu launched an upgraded version of its end - to - end VLA basic model, G0 Plus, in January this year, mainly targeting the generalization of cross - scenario tasks and the optimization of general grasping capabilities, emphasizing the ability of a single set of models to adapt to different forms of robot bodies. In terms of commercial implementation, the company said it had achieved thousands of orders.
After the Billion - yuan Valuation
After the rise of the embodied intelligence track, capital initially focused highly on "robot body" companies. On the one hand, the intuitive abilities such as bipedal walking and jumping demonstrations help to attract market attention first; on the other hand, body companies such as Unitree have accumulated relatively mature technology and commercialization capabilities after a period of development.
In contrast, the more newly - conceptualized embodied "brain" companies were in a catch - up position in terms of capital. In the earlier stages, the embodied brain route was once regarded as "too far - reaching."
Qingzhi Capital invested in Zhipingfang in the angel round in 2023. Zhang Yu, a partner of Qingzhi Capital, recalled in an interview with a reporter from the Science and Technology Innovation Board Daily that the embodied brain route emphasizes end - to - end models and long - term technology investment. In the early stages, most institutions in the industry lacked the corresponding technical understanding ability. "At that time, we analyzed that this route had great potential in the future, but it might face a long period of capital recognition in the short term."
Based on this judgment, Zhang Yu also reminded the founder of a leading embodied intelligence company to control the pace of capital expenditure and arrange the capital cycle well while insisting on technological R & D. As it turned out, the development of the company was indeed as Zhang Yu expected, and the company once had a financing gap of more than a year. "Fortunately, the company had planned its cash flow in advance and was able to smoothly get through the industry's dormant period."
Until early 2025, many investment institutions still maintained a cautious attitude towards the embodied brain. A reporter from the Science and Technology Innovation Board Daily learned that at that time, the valuations of some brain projects still remained in the single - digit hundreds of millions of yuan range, while the valuations of body companies led by Unitree were significantly higher during the same period.
The turning point came in the second half of 2025. Zhang Zhikuan, a director of the investment department of Cornerstone Capital, told a reporter from the Science and Technology Innovation Board Daily, "As multi - modal technology paths such as VLA gradually mature, the generalization ability of robots in cross - scenario tasks such as object grasping and pouring coffee has significantly improved, and brain companies have begun to show stronger technological certainty."
Cornerstone Capital has invested in a series of embodied intelligence projects such as Zhipingfang, Zi Bianliang, and Zhujidongli. Zhang Zhikuan said that although the capabilities of embodied robots are constantly improving and the valuations of projects are constantly rising, the industry as a whole is in the stage of capability iteration and model verification. "It's still far from the time to decide the winner."
He further said that the long - term goal of the embodied brain is to achieve highly general robot intelligence, enabling robots to have understanding and action abilities close to those of humans. However, from the perspective of the current technological maturity, this goal is still far away, and R & D will still be the core task of these projects in the next few years. 'Embodied brain companies rely more on data, computing power, and algorithm iteration, and most of the funds are used for long - term technology accumulation rather than short - term scale expansion.'
A person from an unnamed embodied brain company also told a reporter from the Science and Technology Innovation Board Daily that most of the financing funds would continue to be invested in model training. "Commercialization is of course important, but in the long run, model capabilities determine the upper limit." The person revealed to the reporter that the company had contacted the relevant teams of the Spring Festival Gala but finally prioritized resources for R & D and training systems.
This article is from the WeChat public account "Venture Capital Daily". Authors: Ao Jin, Li Mingming. Republished by 36Kr with permission.