Humanoid robots are rushing to go public.
Embodied Intelligence Booms Unprecedentedly
Since the beginning of this year, "embodied intelligence" has become a hot concept second only to large models. According to Wind data, from the beginning of this year to October 31st, the secondary - market embodied intelligence index has risen by more than 60%, and the robot index has also increased by over 35%.
Let's briefly explain what "embodied intelligence" is. Simply put, embodied intelligence refers to an intelligent system that perceives and acts based on a physical body. That is, artificial intelligence is applied to a body with no fixed form. Currently, it includes various forms such as quadruped robots (robot dogs), wheeled robots, and humanoid robots. Among them, humanoid robots have the highest technical and technological difficulties, and the industry generally believes that humanoid robots have the broadest application space, so they are the most popular direction.
Especially this year, which is known as the first year of mass - production of humanoid robots, has achieved key breakthroughs in policy support and industrial chain maturity, and humanoid robots are starting to move from the laboratory to real - world commercial applications.
However, it remains a mystery and a competitive situation as to who will become the industry leader and go public first.
The R & D of humanoid robots requires continuous investment, and an IPO can provide long - term and stable financial support for technological construction. More importantly, the status of a listed company serves as a reliable trust endorsement, indicating that it has passed the "initial pie - in - the - sky stage" and will have an inherent advantage when competing for large - scale orders and cooperating with leading enterprises. This means that an IPO is no longer just a simple financing activity but a strategic move related to future survival and development rights.
"Showcasing Their Skills" for Listing
To go public, Unitree and Zhipu, which currently lead in valuation, have both taken a series of unusual actions.
First, Unitree, the leader in the humanoid robot market share, is still not satisfied with the current market price and believes it has not reached the bottom.
On July 25th, less than a week after entering the counseling period, Unitree launched its new bipedal humanoid product, Unitree R1. Priced from 39,900 yuan, weighing 25 kg, with 26 joints, and integrating a voice + image multi - modal large model, R1 has further lowered the price anchor of humanoid robots.
Unitree's wheeled dual - arm robot G1 - D
On November 13th, Unitree launched a new full - stack data collection and training solution and introduced the wheeled dual - arm robot G1 - D for the first time. It has 19 degrees of freedom, a vertical working space of 0 - 2 m, binocular + wrist high - definition cameras, and an optional 1.5 m/s mobile chassis. It supports full - process data collection, annotation, asset management, and distributed training, and enters the "data - model - deployment" closed - loop.
Some media voices believe that Unitree only has an advantage in price and strong hardware capabilities but lacks long - term competitiveness. But is this really the case?
Unitree Technology's business path is one of gradual expansion and breakthrough with blockbuster products. Relying on the large - scale sales and profitability of its quadruped robots, Unitree has accumulated experience in motor and reducer technology and global channels, and is gradually expanding from the relatively less technically challenging quadruped balance to bipedal humanoid robots.
Financially, Unitree disclosed that it has been profitable for five consecutive years since 2020. In 2023, its revenue reached 1.84 billion yuan, with a net profit of 210 million yuan, and its net profit has been over 300 million yuan in recent years. This level can outperform some listed companies, indicating that Unitree's commercialization path has been initially successful and it is not a startup relying on financing to survive.
Previously, the market had been speculating that Zhipu would list on the Hong Kong Stock Exchange, and many netizens were ready to subscribe for Zhipu's shares in the Hong Kong IPO. Unexpectedly, in July, Zhipu achieved listing on the A - share market by "acquiring Shangwei New Materials", becoming the "first A - share humanoid robot stock". It even created a 15 - fold increase in stock price, rising from 7.78 yuan per share to a maximum of 132.1 yuan per share. It's no exaggeration to say that at least half of the stock market's enthusiasm for robots this year comes from Shangwei New Materials.
According to Qichacha data, Zhipu has intensively invested in nearly 30 startups within a year, covering core components (such as Fuxing Motor and Qianjue Robot), AI brains (such as Lingchu Intelligence and Xingyuanzhi Robot), and scenario implementation (such as Feikuo Technology).
According to Tencent Finance, to support this investment layout, Zhipu also jointly established a multi - hundred - million - yuan embodied intelligence industry fund with Hillhouse and launched the "Zhipu A Plan" with the goal of incubating over 50 early - stage projects.
Obviously, Zhipu's own financing ability is not enough to support this large - scale layout. After its latest two rounds of financing, its valuation exceeded 15 billion yuan, but its revenue last year was only 100 million yuan. With a high valuation and high losses, Zhipu currently needs two key capabilities: one is financing ability to find a continuous source of funds; the other is mass - production ability, as only mass production can bring in revenue. Therefore, being the first to list and seize the market concept becomes particularly important.
Interestingly, Zhipu seems to have been avoiding the issue of "back - door listing", but a series of its actions make people think otherwise. For example, recently, Peng Zhihui, a co - founder of Zhipu, led a team to independently carry out business at "Zhipu Shangwei". Zhipu has repeatedly stated that there is no business correlation and the two sides operate independently. When interviewed, the securities department of Shangwei New Materials said it had no information about the development of robot business.
After such a back - and - forth, most netizens are probably confused. So on November 11th, Shangwei New Materials issued an announcement stating that the recently launched embodied intelligence robot business is in the development stage, and the company and its affiliated parties independently carry out the embodied intelligence robot business.
It seems to have clarified the matter, but also not really. The author believes that currently, it cannot be simply assumed that Shangwei New Materials is equivalent to Zhipu. Since they are supposed to operate independently, it seems more like an independent brand operation. Currently, Zhipu can either attack or defend. If the market environment is hot, it can achieve an "A + H" listing; if the environment is not good, it can just manage Shangwei New Materials well.
Of course, listing on the Hong Kong Stock Exchange is also a good option. The Hong Kong Stock Exchange's Rule 18C sets two channels for robot companies to list: one is a valuation of at least HK$4 billion plus commercial revenue of at least HK$250 million; the other is a valuation of at least HK$8 billion when the commercial revenue is very low. Under this logic, increasing the valuation and revenue figures has become the main condition for robot companies to list on the Hong Kong Stock Exchange.
Therefore, we can see various embodied intelligence enterprises with the title of "the first" listing on the Hong Kong Stock Exchange. For example, Woan Robot launched an IPO to "strive to be the first AI - embodied household robot stock" and build an ecosystem centered around intelligent household robot products. Geek+ successfully listed on the Hong Kong Stock Exchange, raising HK$2.7 billion and becoming the "world's first AMR warehousing robot stock". Yunji Technology, a hotel service robot company, also successfully listed and became the "first robot service intelligent agent stock".
Some may question the practical significance of competing for the title of "the first stock". After all, it ultimately depends on the product market.
However, the author believes that on the premise that most companies currently do not have large - scale mass - production capabilities, early - stage layout to lock in capital and build one's own industrial ecosystem means winning the first - mover advantage. If startups do not seize the opportunity now, they will have no chance when the industry truly starts mass - production, which has been fully verified in the new energy vehicle industry.
References:
Unitree, Zhipu, and Leju Compete for IPO. Source: 21st Century Business Herald
Embodied Intelligence Stumbles at the Starting Point. Source: Caijing Magazine
This article is from the WeChat official account "Chief Innovation Review Bureau", author: Zuo Jing Guantian, published by 36Kr with authorization.