China humanoid robot forecast doubles to 50,000 units

Morgan Stanley has once again raised its forecast for China humanoid robot shipments, now predicting 50,000 units will ship in 2026 , a figure that has doubled twice in just six months. The investment bank points to a rapid shift from trade-show demonstrations to real-world commercial use in factories, retail, and service sectors.

The numbers tell a striking story. At the start of the year, Morgan Stanley forecast just 14,000 units. That jumped to 28,000 in January, and now stands at 50,000. The bank attributes this acceleration to a faster-than-expected transition from prototypes to paying customers. Commercial adoption, government backing, and supply-chain readiness are converging, according to equity analyst Sheng Zhong. “Commercial verification, policy support, and supply-chain feedback point to faster humanoid adoption in China,” he said.

The market value behind those shipments is equally ambitious. Morgan Stanley estimates China’s humanoid market will hit $2 billion this year, climbing to $15 billion by 2030. By then, annual shipments could reach 446,000 units, up from an earlier estimate of 262,000. These figures exclude in-house prototypes and self-use robots, focusing only on external sales. Globally, the bank expects the humanoid market to expand from roughly $3 billion in 2025 to $28 billion by 2030, a near-vertical climb from last year’s estimated 13,000 units shipped worldwide.

Deployments are moving beyond spectacle. Chinese manufacturers are racing to scale production, placing humanoids in factories, convenience stores, and restaurants. EV maker Xpeng is among those planning mass production by year-end. Morgan Stanley’s supply-chain checks confirm this trend, citing factories, logistics hubs, unmanned shops, and interactive services as early proving grounds. “If you go to any Chinese factory right now, there’s more automation and robotics deployed than anywhere else in the world,” said Joe Ngai, McKinsey’s Greater China chairman. He called humanoids a possible “next big frontier” for investors.

The ambitions extend beyond industrial settings. BYD wants humanoids on its showroom floors to help sell cars within a year or two, its second-in-command told Business Insider. The pitch is shifting from purely operational to customer-facing.

Beijing’s policy support is a major catalyst. The government has made “embodied AI” , intelligence embedded in physical machines , a priority for the next five years. It has directed local authorities to provide cheap land and office space to startups and pushed banks to offer favorable loans. That backing shows in market share: Chinese firms shipped over 80% of the world’s humanoids last year, taking the top five spots by volume, according to Omdia. Unitree led, while America’s Figure AI ranked seventh and Tesla ninth. Tesla does not plan public Optimus sales until late 2027.

For investors, Morgan Stanley recommends a parts supplier over a robot maker. It named Shanghai-listed Leaderdrive a major winner, raising its 12-month target to 464 yuan ($68) from 269 yuan. The firm sells precision components to humanoid builders like Ubtech and Galbot and could hold 40% of the global market this year. It’s a classic picks-and-shovels bet: when dozens of robot brands compete, suppliers of joints, gears, and hands often win regardless. Leaderdrive could retain about a quarter of that market long-term.

The momentum is also driving robotics firms toward public markets. A wave of listings is following the money, with companies like Seer Intelligent now trading in Hong Kong. Seer earns 18% of its sales abroad across more than 65 countries, a rare sign of international reach. The two largest makers, Unitree and AgiBot, are preparing listings that could value them together at around $13 billion.

Geopolitical tensions pose the biggest risk, according to Chinese robot executives. Washington has grown alarmed at China’s AI progress and the prospect of global dependence on Chinese machines. That makes overseas expansion harder. Chinese firms are spreading across markets and emphasizing local compliance to reduce single-country shocks. The robots may be ready before the politics are. Analysts frame it as more than a hardware race: if Washington chases capability benchmarks alone, it may lead on invention but lose on where and how AI gets used. China is selling deployment, not just demos.

The forecast itself warrants caution. A number that doubles twice in six months can move the other way just as fast. Momentum is not the same as durable demand. The math invites skepticism, too. More than 150 firms are chasing a home market that delivered only about 14,000 robots last year. Most will not survive. A shake-out looks likely, and winners will be those that solve commercial deployment, not those with the best stage show. One recent market read found 150 firms chasing buyers who are mostly unimpressed, with only a minority satisfied by the robots on offer. Most deployments still sit in controlled factories where work is predictable. Homes and busy streets remain hard.

Still, the direction is clear. China has made humanoids a national project, and the money, policy, and supply chain are aligning behind that bet. Whether 50,000 ship this year or not, the country is building the machines and the market faster than anyone else. The open question is who ends up buying them.