Humanoid Robot Shipments Jump 272% as AgiBot Overtakes Unitree in China-Dominated Market
China's hold on humanoid robotics now reads as a supply-chain story as much as a research story. Global humanoid robot shipments reached 19,100 units in the first half of 2026, up 272% from 5,100 units in the same period a year earlier, and Chinese manufacturers produced more than 97% of them, according to Smart Analytics Global (SAG), a California-based research firm. The clearest shift inside those numbers is at the top: Shanghai-based AgiBot has overtaken Hangzhou rival Unitree as the world's largest humanoid vendor.
The shipment data reframes the race as a Chinese cost-curve play. Production leadership is real, but unit counts alone will not decide whether the projected 60,000-unit 2026 market converts into durable revenue. What will matter is unit economics, the value robots deliver in actual enterprise deployments, and how Western competitors answer a supply-side concentration with no modern precedent in robotics.
AgiBot Takes the Lead in Humanoid Robot Shipments
AgiBot shipped roughly 8,400 units in the half, a 44% share of the world market. Unitree fell to second place at about 5,900 units and 31%, even though its own volume grew 170% year over year. Combined, the two companies control about three-quarters of every humanoid shipped globally.
| Metric | AgiBot | Unitree |
|---|---|---|
| H1 2026 shipments | ~8,400 units | ~5,900 units |
| Global market share | 44% | 31% |
| Shipment growth year over year | not disclosed | +170% |
| Reported production | 15,000 units by early June | ~11,000 units by March |
The gap between production and shipment counts tells part of the story. AgiBot said in early June that 15,000 units had come off its production line, while Unitree had completed roughly 11,000 units by March, according to MIR DATABANK figures. Neither total is matched by the half-year shipment numbers, which means some of that output is still sitting in inventory or waiting for deployment.
Forecasts assume the momentum continues. SAG projects full-year humanoid robot shipments at roughly 60,000 units and expects the total to climb toward half a million by 2030. TrendForce's separate outlook puts 2026 output growth at about 94%, with AgiBot and Unitree together expected to take nearly 80% of all shipments, a projection the firm bases on the pair's monetization capabilities and mass-production progress.
The Supply Chain Behind the 97% Share
China's dominance is not export-driven. Roughly 85% of demand in the first half came from domestic Chinese buyers, and industrial and commercial settings accounted for more than 70% of all shipments in the period. These are workplace machines tied to deployment budgets, not showroom pieces.
Form factors follow the same commercial logic. AgiBot runs three lines: full-size bipedal A-series robots, half-size bipedal X-series models, and wheeled G-series units. Unitree's flagship G1 is a half-size biped, Galbot sells wheeled machines, and UBTECH pushes its Walker series and U-series toward full-size, ultra-bionic designs.
For enterprise buyers outside China, the concentration is a procurement issue as much as a pricing one. The price war gives buyers leverage: hardware keeps getting cheaper, while the durable value sits in the software, integration, and service layer around each machine. Procurement teams should model the total cost of deployment rather than the sticker price, and treat single-country supply exposure as a risk to price into the decision.
Volume Versus Deployment Value
Leadership by shipment count is not the same as leadership by revenue, and the sector currently splits along those lines. AgiBot leads overall embodied-robot volume, Unitree leads on affordable hardware and pure humanoid shipments, UBTECH leads on audited humanoid revenue, and Figure AI leads on general-purpose autonomy. Each vendor is betting on a different definition of winning.
The pricing pressure behind that volume is intense. Unitree has brought a humanoid to market at the $4,290 mark as part of an aggressive cost war, which compresses margin across the whole category. At those prices, profit depends on component supply, manufacturing scale, and service revenue rather than the sticker price.
Unitree's own numbers illustrate the tension. The company had completed roughly 11,000 units by March yet shipped about 5,900 in the entire half, a spread that points to inventory build or machines still waiting for deployment. Volume built on discounted hardware only becomes durable revenue if those units reach paying deployments quickly.
There is a deployment reality check underneath the production boom. Chinese makers have moved to the front of humanoid AI benchmarks, so the software gap that once protected Western vendors is closing, yet the question of whether these machines earn their keep in factories and commercial settings remains open. The 60,000-unit forecast is a supply forecast; the demand side still has to prove it can absorb that output at prices that leave margin intact.
The IPO Wave and the Western Response
Both leaders are heading to public markets. Unitree's Shanghai listing drew nearly 10 million retail orders and ran more than 5,500 times oversubscribed, which would make it mainland China's first publicly traded humanoid robot maker, with the deal valued around $6.2 billion. AgiBot is preparing its own listing as the physical AI wave draws robotics startups, major tech companies, and automakers into the sector.
Public-market pressure cuts both ways. A listing forces quarterly revenue discipline on companies that have so far competed on volume and price, and it gives investors a direct view of whether humanoid sales actually produce margin. For Western vendors the math is uncomfortable: competing against $4,290 hardware built on Chinese supply chains means the fight will be settled on software, autonomy, and enterprise services rather than manufacturing cost.
Figure AI's position as the general-purpose autonomy leader shows where the non-Chinese field is headed. With 97% of hardware concentrated in Chinese factories, Western players are effectively ceding the cost curve and betting that intelligence, integration, and deployment know-how command a premium that volume alone cannot match. That bet is the mirror image of the Chinese strategy, and both cannot hold at the same price point.
For buyers and investors, shipment leadership is the wrong lens. The question is whether the 60,000 robots projected for 2026, and the half-million forecast for 2030, are deployed in settings where they earn their keep, and which vendors convert that deployment into margin before the price war compresses it further. AgiBot holds the volume lead, Unitree holds the pricing and IPO momentum, and the rest of the field competes on software. The vendor worth watching is the one that closes the gap between production output and paying deployment first.
Why This Matters
The 97% concentration turns humanoid robotics into the same kind of supply-chain question the semiconductor industry already faces: whoever controls manufacturing sets the floor price for everyone else. Enterprises evaluating humanoids should treat hardware as a commodity that keeps getting cheaper and budget for software and integration value instead, while investors should read humanoid robot shipments as a production metric rather than a revenue one.
Related Articles
- Unitree IPO Draws 5,526x Retail Demand in Shanghai's First Humanoid Robot Listing
- Unitree IPO Prices at 219x Earnings as Humanoid Robotics Gets a Daily Market Test [Update]
- Humanoid Robotics IPO Hype Masks Unitree Profit Crisis
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.