AgiBot overtakes Unitree for the first time: a new balance of power in the humanoid robot market
The first half of 2026 was marked by a historic leadership change in the humanoid robot segment. According to my analysis of Smart Analytics Global (SAG) data, China's AgiBot has for the first time surpassed Unitree, taking a dominant position in the global market. AgiBot's shipment volume reached approximately 8,400 units, corresponding to a 44% market share, while Unitree, with around 5,900 devices, dropped to second place with 31%.
In total, from January to June, manufacturers shipped about 19,100 humanoids — a more than threefold increase compared to the same period last year. Notably, Chinese companies have virtually monopolized the industry, accounting for 97% of all global shipments.
Shift from experiments to industry
The key takeaway I draw from the report is a radical change in demand structure. While a year ago humanoids were used primarily for research purposes, now more than 70% of shipments are for industrial and commercial tasks (versus ~50% in the first half of 2025). Robots are increasingly being integrated into production lines and warehouse operations that require repetitive physical actions.
However, it is important to understand: such high shipment rates do not always reflect actual revenue. A significant portion of shipments consists of platforms for pilot projects and software development, indicating an early stage of monetization rather than mass adoption.
Strategic differences among leaders
The competition between AgiBot and Unitree is built on different approaches. Unitree has earned a reputation through affordable solutions for researchers and developers, while AgiBot has focused on expanding its industrial portfolio. The gap of 2,500 units is not just a number, but a signal that the market prefers practical scenarios over demonstration models.
The geography of shipments is also expanding: Chinese manufacturers are actively increasing exports, challenging American and European competitors on their own turf.
Forecasts and growth barriers
SAG expects global shipments to approach 60,000 units with revenue of about $1.6 billion by the end of 2026, and the market to nearly double to $3 billion by 2027. By 2030, the volume could reach 500,000 robots. But, in my view, the key factor will be not production capacity, but technological maturity. Among the main obstacles are a shortage of real-world training data, limited AI models, reliability, manipulation precision, safety, and cost.
Multimodal VLA and world models are evolving rapidly, but are still in their infancy. The industry stands on the threshold of transitioning from impressive demonstrations to reliable operation in the real world — and it is this barrier that will determine future leaders. Additionally, geopolitical risks, including recent U.S. restrictions on imports of foreign robots, could seriously hamper the expansion of Chinese companies.
As for home robots, in my assessment, universal humanoids for household use will remain a niche product at least until the early 2030s due to the complexity and unpredictability of residential environments.
My comment: The leadership change is a natural outcome of market evolution. AgiBot won not through revolutionary technology, but through precisely hitting industrial demand. However, investors should be cautious: current shipment volumes may create an illusion of maturity, while real profitability and system reliability remain in question. The next 12–18 months will be decisive in testing these ambitions in practice.