China's humanoid robotics sector is establishing early global dominance, with companies like Unitree and Agibot leading in shipments and rapid technological iteration. This acceleration is propelled by the nation's advanced manufacturing ecosystem, supportive government policies under initiatives like "Made in China 2025," and significant private capital investment, positioning the country to address looming labour shortages.
The industry's shift from demonstration-driven showcases to operations-driven adoption is now central to its growth, as companies focus on proving value in real-world environments such as factories and warehouses.
Hardware and Manufacturing Edge
Analysts attribute China's lead to a superior hardware supply chain, largely built through its electric vehicle sector, and the world's strongest manufacturing base. This allows Chinese companies to iterate and release new models far faster than Western competitors. Selina Xu, a China and AI policy lead, stated that leading Chinese player Unitree shipped roughly 36 times more units in 2025 than US rivals Figure and Tesla combined.
Global humanoid robot shipments totalled just 13,317 units in 2025, a tiny base for an industry projected to reach 2.6 million units by 2035. The top shippers were China's Agibot and Unitree, followed by UBTech, Leju Robotics, Engine AI, and Fourier Intelligence.
From Demos to Deployments
The market focus is transitioning from "demo-driven excitement" to "operations-driven adoption," according to Yuli Zhao, chief strategy officer at Galbot. "More customers are asking: Can the robot run stably in real environments and actually take work off people’s plates?" Zhao told TechCrunch. This practical demand is strengthened in China by policy encouraging automation upgrades.
Chinese firms are securing substantial funding to fuel this growth. Unitree was valued at around $3 billion after a Series C round, while Galbot has raised over $300 million, pushing its valuation to an estimated $3 billion.
Software Challenges and Global Competition
While hardware advances quickly, the AI "brains" for humanoids remain nascent. The industry bets on vision-language-action models, but faces a critical data scarcity problem for training, as companies cannot simply scrape the internet like large language models do. Most rely on simulation environments for synthetic data, though real-world collection is essential.
Nvidia currently leads in end-to-end humanoid software, with most Chinese startups using its Orin chips, though domestic alternatives are in development. Safety remains a major hurdle, with regulations expected to increase as the industry matures.
Competition is not limited to a US-China race. Japan's robotics ecosystem, a long-time pioneer, is targeting mass production by 2027, leveraging precision engineering and a cultural affinity for robots, particularly in eldercare. Hyundai's Boston Dynamics also plans to produce up to 30,000 Atlas humanoids annually in the US by 2028.
Future Trajectory and Scale Advantage
Early commercial momentum is expected in industrial manufacturing, warehouse logistics, and retail, where tasks are repetitive and processes clear. A December TrendForce report notes China is already targeting a mix of affordable mass-market models and high-end applications across various sectors.
"China’s leadership is best understood as a speed-to-scale advantage," Zhao concluded. "The ecosystem here compresses the entire cycle — R&D, supply chain, manufacturing, integration, and customer deployment — into a very tight loop... allowing iteration at a pace that’s difficult to match elsewhere."