TL; DR
China will supply 90 percent of the world’s humanoid robots in 2025, and there are more than 150 companies in the sector, but only 23 percent of businesses surveyed are satisfied with existing products. Morgan Stanley is warning of an upheaval as billion-dollar IPOs collide with two-hour battery lives and a market that shipped just 14,000 units last year.
There are more than 150 humanoid robot companies in China. It will supply nearly 90 percent of the world’s humanoid robots in 2025. Its two largest manufacturers, Unitree and AgiBot, are preparing initial public offerings with a combined value of $13 billion. Morgan Stanley it has doubled its supply forecast for the Chinese market this year to 28,000 units, a 133 percent increase over 2025. However, when Morgan Stanley surveyed companies that should buy these robots, only 23 percent said they were satisfied with existing products. Battery life lasts two to three hours per charge. Most deployments are limited to exhibitions, exhibition halls and Spring Festival galas where robots perform kung fu exercises for television cameras. Technology has arrived. Customers don’t have it. China’s humanoid robot industry is the world’s most capitalized, most productive and most populous robotics sector, and it’s headed for a reckoning its government has already warned it will be.
Warning
At the end of 2025, China’s National Development and Reform Commission made a rare public statement about the humanoid robot sector. Spokesman Li Chao noted that the number of companies has exceeded 150 and is still growing, with more than half being startups or cross-industry participants. The NDRC warned of redundant products, duplicative investments and squeezed space for genuine research and development. Language measured. There was no effect. Beijing’s top economic planning agency told the market there was a bubble in the industry, which it identified as one of ten priority sectors in the 15th Five-Year Plan and is backed by a trillion-yuan state fund.
China’s smartphone supply chain has already begun to focus on the production of humanoid robotsWith companies like Foxconn supplier Lingyi iTech assembling iPhones targeting 500,000 humanoid devices by 2030. The production infrastructure is real. The component ecosystem is deep. The problem is that the robots in production are not yet generating the revenue they are valued at. Unitree, which filed for a $608 million IPO on Shanghai’s STAR market, saw revenue from humanoid robots surpass its quadruped robot business for the first time in 2025, but the company’s overall scale remains modest compared to its targeted $7 billion valuation. AgiBot, which is aiming for a $6 billion listing in Hong Kong, is in a similar position: significant technological capabilities, significant government support and an unrealized commercial market on the scale of IPO price demands.
Gap
A Morgan Stanley survey led by China industry analyst Sheng Zhong said 62 percent of Chinese companies will adopt humanoid robots within three years. However, this desire collided with a number of practical limitations that the industry has not resolved. The 23 percent satisfaction rate reflects shortcomings in flexibility, functionality and price. Ninety-two percent of respondents said that before mass adoption of robots is possible, the price needs to fall below 200,000 renminbi, about $28,000. Only about 10 percent of companies surveyed are currently evaluating or implementing pilot projects. The demand is theoretically there. In practice, robots are too expensive, too limited in capabilities, and too short in battery life to justify the investment for most industrial applications.
UBTech, one of the biggest players in the sector, has offered $18 million to hire a chief artificial intelligence scientist.a salary that reflects both the intensity of the talent war and an acknowledgment that the remaining engineering challenges are significant. UBTech’s latest industrial humanoid, the Walker S2, entered mass production in early 2026 with orders exceeding 800 million yuan, and the company is building a factory in Beijing to produce 10,000 units a year by the end of 2026. But production capacity and commercial demand are different things. Morgan Stanley’s Zhong sees 2026 as “a critical year as humanoid integrators strive to achieve commercialization and build their ecosystems,He noted that production is likely to be significantly larger than sales because major players are building robots in-house for training and testing rather than shipping them to paying customers.
The play
In April, Lightning, a humanoid robot developed by Chinese smartphone maker Honor, won the Beijing E-Town Half Marathon in 50 minutes and 26 seconds.broke the human world record by nearly seven minutes. More than a hundred robots competed. The event was covered worldwide. The winning team’s engineer said the achievement allowed the technology to transition to structural reliability and cooling, which will eventually benefit industrial applications. Robotics experts were less certain. Skills demonstrated during sustained bipedal locomotion on a flat surface during a half-marathon do not translate into manual dexterity, real-world perception, and adaptive problem-solving required for factory work, logistics, or service applications on which industry business plans depend.
The gap between spectacle and substance defines China’s humanoid robot moment. Spring Festival Gala performances, marathon records and viral videos of robots doing backflips attract attention that attracts capital. Capital funds the next stage of development. Development creates more effective demonstrations. But the turnover does not generate income on the scale necessary to justify the valuations set. China’s industrial model has historically excelled at commercializing technology faster and cheaper than any Western economymaking solar panels, electric vehicles and batteries a globally dominant export industry within a decade. The question is whether humanoid robots will follow this pattern or represent a category where the gap between display and deployment is wider than manufacturing advantage.
Competition
China’s dominance in the supply of humanoid robots has not gone unnoticed. Boston Dynamics began commercial production of the electric Atlas robot in January 2026 and announced plans to deploy tens of thousands of units at Hyundai Motor Group factories with its manufacturing facility near Savannah, Georgia, aiming for 30,000 units a year by 2028. Stockpiling in 2025 despite shipping some of the volume managed by Chinese companies. With Tesla Optimus performing key tasks in its factories, Elon Musk is predicting mass production and a price point between $20,000 and $30,000, although the robot, by Musk’s own admission, “not used financially.” The Pentagon has signed a $24 million contract with Foundation Future Industries for humanoid robot soldiers tested in Ukraine.The opening of the military market, which Chinese companies cannot enter, but which confirms the strategic importance that governments attach to the technology.
Price dynamics are in favor of China. Unitree’s H2 is priced under $30,000. Another Chinese manufacturer, Kepler, is targeting the same range. The large number of Chinese humanoid robots on display at CES 2026 and their aggressive pricing made it clear that the supply economy is already competitive. The question is whether the demand at these price points is high enough to sustain an industry where 150 companies compete.
Calculation
Zhong’s concussion prediction is not a minority view. The NDRC’s warning, Morgan Stanley’s satisfaction data, Unitree’s IPO inspection just twelve days after the STAR Market application was accepted, and simple arithmetic of 150 companies chasing a market expected to supply about 14,000 units in China in 2025 all point in the same direction. The companies that survive will be the ones that solve the commercialization problem: identifying repeatable, scalable use cases where the economics of a humanoid robot outperform the alternatives, whether those alternatives are purpose-built industrial weapons, wheeled platforms, or human workers. Companies that don’t will be burning through their finances by producing impressive machines that no one needs to buy outside of a trade show.
China’s humanoid robot industry has the manufacturing base, component supply chain, government support and engineering talent to lead the world. It’s a market where it doesn’t exist yet. A trillion yuan sovereign fund and the 15th five-year plan ensure that capital flows will continue. The NDRC warning ensures Beijing is watching how it flows. Between billion-dollar IPOs and a 23 percent satisfaction rate, marathon records and two-hour battery life, the answer is whether China’s humanoid robot boom will produce the next big Chinese export industry or the most expensive collection of trade show displays the tech sector has ever funded. Robots can run a half marathon faster than any human alive. They can’t work eight-hour shifts yet.
