The United States has moved to restrict foreign-made humanoid and quadruped robots from China, a decision that we think will hit individual companies with heavy American exposure far harder than it will slow the broader growth of China’s robotics sector.
Modest but Real Export Volumes
Chinese Customs’ data offer one of the few concrete proxies for how many robots have reached American shores. Under HS code 84798970, China exported 10,400 units worth RMB717.7 million in the first half of 2026. Of those, 1,840 units worth US$23.3 million went to the United States — about 18% of total export volume, enough to make the US a meaningful destination for Chinese robotics exports.
The figures come with an important caveat: HS 84798970 is a broad customs category that covers machinery and mechanical appliances well beyond humanoid and quadruped robots. We believe the 1,840-units figure should be read as an indication of the overall scale of China’s robotics-related exports, not as an exact count of the robots covered by the US restriction.
At the company level, Unitree is regarded as a significant Chinese exporter to the US among humanoid and quadruped manufacturers, with products sold widely to international customers including American universities and research users. AgiBot is another notable exporter, and UBTECH remains an important Chinese humanoid manufacturer, though scant public information exists on the geographic breakdown of its shipments. US import records have specifically documented AgiBot’s humanoid and quadruped platforms entering the country, confirming that Chinese-made advanced robots are already circulating in the American market.
A Meaningful Revenue Stream — For Some
For a handful of companies, US sales already represent a meaningful revenue stream, though not for the Chinese robotics industry as a whole. Unitree is the clearest example: the company generated roughly RMB1.7 billion in revenue in 2025, with more than 40% coming from overseas markets. The US alone accounted for up to about 18% of revenue, according to data from the company’s recent IPO materials.
That makes the US a meaningful market for Unitree, and losing access could noticeably affect its revenue growth — particularly because the company has been among the most successful Chinese firms at selling relatively low-cost robots overseas. However, we don’t think we should extrapolate Unitree’s US exposure across the industry, however. Many Chinese robotics companies remain overwhelmingly focused on the domestic market, which is itself large and growing rapidly. Customs data reinforce the point: even within the broader HS 84798979 category, the US represented only around 18% of China’s export volume in the first half of 2026, leaving a substantial non-US market.
FCC Action Raises the Stakes for US-Focused Firms
The US Federal Communications Commission’s decision adds new foreign-made humanoid and quadruped robots to its Covered List, meaning new models generally cannot obtain the FCC authorization required to be marketed or sold in the country. The immediate impact will fall hardest on companies that have already built out US distribution channels — Unitree being the clearest case, given that the US can represent a high-teens share of its revenue.
We think there are two reasons the impact could run deeper than the direct revenue loss suggests. First, the US is an important technology and research market: Chinese companies have sold robots to American universities, laboratories and companies that provide not just revenue but application feedback and opportunities to refine products in a sophisticated market. Second, the restriction arrives just as humanoid robotics enters a critical commercialization phase, and losing a large developed-market customer base could slow international commercialization and narrow the range of applications in which Chinese robots are deployed.
Even so, the sector-wide impact is expected to be far smaller than the hit to individual companies. China’s domestic market is already enormous, and Chinese manufacturers hold a substantial lead in shipment volumes. Omdia estimates Chinese companies accounted for the vast majority of the roughly 13,000 humanoid robots shipped globally in 2025, with AgiBot, Unitree and UBTECH among the leading suppliers. Rather than fundamentally altering the trajectory of China’s robotics industry, the ban is more likely to redirect growth from the US toward China and other overseas markets.
Chinese Firms Weigh Their Response
Chinese robotics companies have several straightforward strategies available to offset the setback. The first is to accelerate growth outside the US. China itself is the most obvious alternative: domestic humanoid shipments were already estimated at around 18,000 units in 2025 and are expected to rise substantially in 2026. Companies can also target Europe, Southeast Asia, the Middle East and other markets with currently lighter regulatory barriers — a path particularly feasible for firms such as Unitree that already have international distribution networks.
A second strategy is deeper localization overseas. Over time, Chinese robotics companies could establish local assembly, manufacturing, R&D or distribution operations in non-US markets, easing compliance with local rules and reducing reliance on direct exports from China. We think, however, that simply assembling a Chinese-designed robot abroad would not necessarily circumvent the US restriction if the policy targets the origin and underlying technology of the robotic system — meaningful localization would require a much deeper restructuring of the supply chain.
Third — and perhaps most important — companies can lean further into China’s domestic commercial market. The next phase of the industry is less about demonstrating that a robot can walk and more about proving it can perform economically useful work, and China’s vast manufacturing base offers ample opportunity to test robots in logistics, automotive, electronics, warehousing and other applications.
Finally, Chinese companies can keep strengthening their vertically integrated supply chains. China already has strong capabilities in motors, batteries, actuators, sensors, precision machinery and electronics — an advantage in cost and production scale that could help offset the loss of some Western markets.
China Leads in Scale, US Leads in Intelligence
Broadly, China currently holds a clear advantage in robotics manufacturing scale, hardware supply chains and early commercialization, while the US retains significant strengths in AI, software, computing and venture capital. The gap is especially visible in humanoids: Chinese companies dominated global shipments in 2025, with AgiBot shipping more than 5,100 units, Unitree around 5,500 and UBTECH about 1,000, according to industry data.
That gives Chinese companies an important edge, since robotics is ultimately a hardware-and-manufacturing business as well as an AI business — and China already has the EV, electronics and industrial-automation supply chains needed to produce large numbers of robots and components at relatively low cost. The US, by contrast, leans on a different set of strengths: companies such as Tesla and Figure AI have strong capabilities in AI, foundation models, computing and software, keeping the US highly competitive in the intelligence layer of robotics even as China leads in physical production volume.
We would not equate shipment volume with technological leadership. Humanoid robotics remains at an early stage, and a significant share of current shipments are going into research, demonstrations and pilot projects rather than mature commercial applications. The real competitive test will be whether companies — Chinese or American — can achieve reliable performance and attractive returns on investment in large-scale industrial and commercial deployments.
What to make of Unitree’s upcoming IPO tomorrow
Unitree’s IPO is significant because it provides an important A-share valuation benchmark for embodied AI and humanoid robotics. UBTECH has already been listed in Hong Kong, so Unitree is not China’s first listed humanoid robotics company. But it gives mainland investors direct exposure to one of the sector’s leading companies and could influence valuations for future robotics IPOs and private companies.
For Unitree, I would focus less on the absolute size of the IPO and more on what its valuation and aftermarket performance tell us about investor appetite for robotics. I would watch its valuation, evidence of commercial adoption beyond demonstrations, and whether investor interest spills over into other robotics companies and upstream suppliers.
