From hesitating to invest to lamenting high valuations to diverging paths, the speed of change in investment trends rivals the rapid iteration of humanoid robots. Even as the industry's technical trajectory has yet to converge, investors have reached a preliminary consensus: "Projects focused solely on the physical body are not invested in."

Editor: Di Xintong

"I don't understand, so I dare not invest."

Earlier this year, when the Embodied Learning Society spoke with several investors, a regretful sentiment akin to "beating oneself up" emerged as the primary theme of the conversation. Particularly striking was the investor who had invested in nearly every type of robotics company but shied away from humanoid robots. When asked if they regretted their decision, they responded with silence.

The investor didn't speak, just smiled awkwardly, "Not many people understand it!"

Even the handful of humanoid robot companies that were once "the hottest stars" found themselves on the cold bench of investors a few years ago. One industry investor told the Embodied Learning Society that they had been in contact with a humanoid robot company since its inception. While they were aware of its hardware advantages, they still hesitated to invest. Another industrial investor echoed this sentiment, lamenting that at the time, they didn't know what humanoid robots could do and now it's too late to grab market share.

"Do you know what they can do now?"

"Well, it's not just for dancing."

We cannot judge investors' choices of the past with hindsight. After all, in the two years since the emergence of large models, the competition for model parameters has been fierce. Emerging large models, viewed from any angle, are currently the hottest and most promising targets aligned with long-term investment logic. In contrast, a humanoid robot still spinning in the laboratory lacks immediate appeal and perhaps overpromises, making it difficult to foresee short-term returns and long-term direction. After all, this is an industry that is "always five years behind."

But hesitation for a moment doesn't mean staying on the sidelines forever. Just a few months later, when the Embodied Learning Society followed up with the aforementioned investors, it was discovered that some institutions had started welcoming humanoid robot projects to their meetings. Now, not only do they understand the technology, but they also understand it in greater depth. Even as the company's technical route has yet to converge, investors have reached a preliminary consensus: "Projects focused solely on the physical body are not invested in."

The mantra of mass production in the first year has been chanted for half a year, and the inflection point for humanoid robots seems to be gradually becoming clearer. In mid-2025, the Embodied Learning Society spoke with investors of various backgrounds from different perspectives, using the investor's perspective as an entry point to glimpse the undercurrents hidden beneath the humanoid robot fervor.

"Love the brain, love the upstream, firmly believe but still dare not be overly optimistic" may encapsulate the current investors' mindset.

No Longer Pursuing the Physical Body Alone

From embracing the physical body to lamenting its high cost to disregarding projects solely focused on the physical body, this evolution in investment trends has transpired in just one year.

Data offers the most intuitive reflection. According to incomplete statistics from the New Strategic Industry Research Institute, Chinese humanoid robot companies received over 50 financing rounds in 2024, with a total financing amount exceeding 5 billion yuan. From the perspective of the invested companies, most belonged to physical body enterprises, such as Zhiyuan Robotics and Unitree Robotics, both of which secured single-round financing exceeding 1 billion yuan.

From 2024's perspective, the physical body was sexier and offered more room for imagination. However, due to the projects' early stages, more evaluation dimensions still hinged on the founding team's background. The founder's impressive resume was closely tied to the amount of financing. "At that time, it was more about a betting mentality, so we paid more attention to the team's background and were willing to pay a certain premium," investor Wang Qiang recalled to the Embodied Learning Society.

This wave of "physical body investment fever" continued until shortly after Unitree Robotics appeared on the Spring Festival Gala. Under the spotlight, the valuation of physical body companies soared. Investors still believed in the physical body's story, but under the influx of hot money, valuation became a barrier that deterred most investors from entering the market.

Image source: Unitree Robotics

"The physical body is too expensive" has gradually become the consensus at this stage.

Data shows that there were over 30 financing events in the first quarter of 2025, with a total amount exceeding 4 billion yuan. Among them, financing events exceeding 100 million yuan accounted for 63%, and those exceeding 500 million yuan accounted for 14%. It's worth mentioning that as heavy investment in humanoid robots continued, divergence in investment paths began to emerge. The physical body remained an excellent target, but some investors started paying attention to more segmented development, staging an ecological positioning war.

The mainstream directions at this stage are: focusing on the brain or algorithm; focusing on the upstream. Those who bet on the former respect the essence of technological development more. After all, the humanoid robot is about the "brain" rather than the "form." Its previous inflection point benefited from the technological leap of large models, and the next inflection point is also inseparable from a smarter brain. Therefore, betting on the "brain" obviously has a brighter future. It can also be seen from the first quarter's financing that the number of humanoid robot enterprises with large models as their core is gradually increasing.

The latter pays more attention to business logic. The mantra of mass production in the first year was already raised at the beginning of 2025. The number of humanoid robots rolling off the production line marked the shipment of tens of times the number of core components. According to Goldman Sachs research reports, the global market size of core components for humanoid robots will exceed 500 billion yuan by 2030. This is a business that can see financial boosts in the short term. In the future, as technology matures and costs are spread out, shipments of humanoid robots will see a significant increase. Core components may witness the golden age of the automotive upstream industry, cultivating a "CATL" equivalent.

History often repeats itself. This period is strikingly similar to the investment divergence seen in the new energy vehicle industry. After the financing fever for OEM factories, investors began betting on intelligent systems or upstream components. There is no right or wrong, only different starting points.

On the component side, there are two types of targets. The first type is represented by mature manufacturers, most of which come from the automotive industry's upstream. They possess the foundation for mass production and can form a logic of technology reuse, with strong overall risk-resistance capabilities. The method of separately opening production lines also reduces the cost of entry. However, like physical body companies, these manufacturers still tout high valuation stories.

The second type is startup enterprises with a short history. Since their inception, they have targeted humanoid robots and focused on integrated joint modules, etc. Although these enterprises may not have the accumulated capabilities of mature manufacturers, their valuation is relatively low, and there are also opportunities for development. However, unlike physical body startups, when investors pay attention to these manufacturers, they don't require the founder to come from a well-known laboratory or have a university background. Instead, they prefer entrepreneurs who have emerged from the real industry. Understanding the production line is more attractive than understanding technology.

It's evident that hot money isn't blindly pouring into the humanoid robot industry. Investors are discreetly sorting out more segmented investment directions that the spotlight can't reach. In an uncertain industry, they're finding a sense of certainty to rely on. When discussing their current investment mindset, most investors hold a rational and optimistic attitude. There will definitely be valleys, and bubbles will definitely burst, but the trend is irreversible.

This isn't being friends with time; it's being comrades-in-arms with the era.

"Physical Body +" is the New Sexy

Commercialization Must Be Discussed

As segmentation becomes a crucial trend in the divergence of investor paths, the simple story of the physical body appears somewhat insufficient, and the concept of "physical body +" emerges as the times demand.

The essence of this investment is akin to the underlying logic of the aforementioned "brain investment," both of which pay more attention to technological breakthroughs on the physical body side rather than simply "popping out" a humanoid robot. The large model capabilities of Galaxy General, the "one brain, multiple forms" of Xinghai Map, and the frequent large financing rounds secured by multiple physical body enterprises that have launched dexterous hands are all significant cases that prove the evolution of investment trends.

Without demonstrating technological differentiation, it's impossible to tell a complete story of the physical body, which has become the main tone of this investment trend.

"It's a bit like buy one, get one free. The physical body tells the current story, and what comes after the plus sign is about future sustainable development. It's necessary to give investors reassurance and also provide them with a long-term strategy," industrial analyst Meng Meng told the Embodied Learning Society.

It's worth mentioning that what comes after the plus sign isn't necessarily the iterative capabilities of software or hardware but also the presence of commercialization. Such as "physical body + scenario" and "physical body + case."

"I pay more attention to the ideas or advantages of cutting into scenarios," investor Dai Ji told the Embodied Learning Society. In Dai Ji's view, the emergence of technological inflection points is the result of the entire industry, not solely relying on a single company. Under the influence of technological convergence and the cycle, those with stronger commercialization capabilities or scenario advantages have more investment value. After all, most companies currently follow the business logic of "looking for nails with a hammer," and future initiative rights are worth considering.

For instance, the application of factory scenarios is gradually becoming homogeneous, and the implementation of inspections and transportation is not uncommon; the imagination space for scientific research scenarios is relatively small. Although it can support order data in the short term, it's difficult to become the main revenue stream in the long run. Therefore, those who can dig into scenario incisions and humanoid robots that grow out of scenarios may be more capable of completing early commercialization and monetization.

From the enterprise perspective, rolling out scenario verification and mass delivery is no longer novel, and companies have also started being strongly tied to scenarios, such as cultural tourism, commercial use, and retail. There are also companies that choose to expand horizontally, verifying in dozens of scenarios, and ultimately finding the golden balance point that is more suitable for short-term boosting and long-term development space.

Taking a certain case as an example, the company chose the logic of "small b to c," starting from the small b-end of commercial scenarios. After accumulating initial scenario experience, it gradually expanded to the c-end, allowing it to stand out among a group of physical body enterprises that bet on industrial scenarios.

"Physical body + case" is slightly different from scenario. This logic is more pragmatic, focusing on whether there are sufficient actual commercialization cases to support it at this stage. For example, cooperation with large enterprises or new cases that open up new scenarios.

Whether it's scenario or case, it indicates that investors pay more attention to more pragmatic commercialization capabilities. Or to delve deeper, they're concerned about whether humanoid robot enterprises are walking on the correct commercialization path. Cool technology is certainly appealing, but the outcome of Boston Dynamics is evident to all.

Written at the End: Cautiously Optimistic

When communicating with all investors, it's always impossible to avoid discussing their current mindset.

After all, investors are facing an innovative industry of far-reaching significance, and its investment value far exceeds financial returns. However, its high-tech attributes also lead to the risk of "drawing water with a bamboo basket" if one is not careful. From autonomous driving to AI chips, the periodic fluctuation pattern of hard technology investment has repeatedly confirmed that during the trough period of the technology maturity curve, slight deviations in judgment may lead to the dilemma of "imbalance between innovation investment and commercial returns."

This collective anxiety condenses into a delicate contradiction in the hearts of investors - they firmly believe that what they hold is the "ticket to the future industrial kingdom" of the era, but they're also alert to whether this ticket bears the invisible watermark of "technological bubbles."

Amidst the dual variations of era opportunities and technological risks, "cautious optimism" has emerged as the consensus within the industry. It involves prudently evaluating the commercialization conversion rate of technological pathways while optimistically embracing the historical inevitability of the intelligent equipment revolution. At its core, this mindset reflects professional investors' profound understanding of the "dialectics of high-tech industrial development." It signifies embracing change while maintaining a rational awe, and striving to find the fulcrum of certainty in industrialization amidst uncertainty.