Chinese Firms Dominate Global Humanoid Robot Revenue Ranking, Frost & Sullivan Says(Yicai) Sept. 29 -- Chinese companies took all five top spots in the global ranking by humanoid robot revenue from 2025 through the first half of this year, according to Frost & Sullivan.
The five companies together accounted for more than 50 percent of the global market, the consultancy said in its 2026 humanoid robot market report. Global sales reached 36,000 units in the first half, generating USD540 million in revenue, already exceeding the USD500 million recorded for all of last year.
Ubtech Robotics, the Chinese firm behind the Walker series, ranked first worldwide last year, with revenue of CNY870 million (USD130 million), shipments of nearly 13,840 units, and an almost 25 percent revenue share, according to the report. The Shenzhen-based company retained the top spot in the first half, when revenue reached CNY620 million, and sales exceeded 16,120 units.
China's Agibot, Unitree Robotics, Galbot, and Leju Robotics Technology followed Ubtech in the ranking. Overseas companies, including Tesla and Figure AI, remain largely at the prototype iteration and internal-use stages, the report said.
Frost & Sullivan divides humanoid robot applications into four categories: entertainment performances, scientific research and education, data collection, and production operations.
Production operations recorded the strongest revenue growth among the four application areas. Global sales of humanoid robots used in production operations totaled 700 units last year, generating USD47 million in revenue. Sales in the segment climbed to 1,200 units in the first half of this year, while revenue reached USD78 million, up 66 percent from the total for all of 2025.
Meanwhile, the entertainment performance market -- the largest segment -- remained broadly stable, while the scientific research and education market declined slightly. Data collection applications were also largely unchanged in scale and were mainly used to support embodied artificial intelligence model training.
Editor: Emmi Laine
