Picture the cost of a worker who never sleeps, never asks for a raise, and never calls in sick. Now picture that worker costing less than a mid-spec Toyota Camry. That's not a thought experiment anymore. In Hangzhou, a company called Unitree will sell you a walking, balancing, two-armed humanoid robot — the G1 — starting around $16,000. Its smaller cousin, the R1, goes for about $5,900.
Let that price sink in, because it's the whole story. For most of the last decade, "humanoid robot" meant a $250,000 lab curiosity or a slick Boston Dynamics video that never went on sale. China just collapsed that price by more than 90%, and in doing so it quietly took over an industry the West assumed it would lead.
If you've been watching robot clips rack up views on LinkedIn and wondering whether there's actual money here — there is. But it's not always where you'd expect, and the people who'll profit most aren't the ones building the robots. Let me walk you through who's cashing in, because some of those people could be you.
The number that reframes everything
Here's the stat that should stop you cold: according to Morgan Stanley, Chinese makers account for roughly 90% of the 13,000 to 16,000 humanoid robots shipped globally, while the US and Japan are mostly still at the prototype stage. Nine out of ten humanoids walking around a factory right now came from China.
This isn't a forecast. This is who already won the first round, while most of the Western press was still arguing about whether Tesla's Optimus could fold a shirt on stage without a human puppeteering it from backstage.
And the demand behind that number is real, not hype-cycle vapor. Jensen Huang, the CEO of Nvidia — the company selling the chips that make all of this possible — stood on stage at CES 2026 and called it "the ChatGPT moment for robotics." He's been telling investors that humanoid labor is a $40 trillion total addressable market. Goldman Sachs, far more conservative, projects the humanoid market hits $38 billion by 2035 — a sixfold increase over its own earlier estimate, revised up after AI progress "surprised the team the most." Morgan Stanley goes further still: $5 trillion by 2050, with close to a billion robots in the world, 90% of them doing industrial and commercial work.
When the chip seller, the most bullish bank, and the most cautious bank all point the same direction, the trend itself stops being the question. The only question left is who captures the money.
Who's actually making money right now
It's tempting to assume the money is in building the most advanced robot. It isn't — not for most people. Here's where the cash is actually changing hands today.
The robot makers themselves — and their shareholders. Unitree, founded by Wang Xingxing in 2016, isn't a science project burning venture money and posting demos. In 2025 its revenue grew 335% year over year to ¥1.708 billion (about $235 million). In March 2026 it filed for a $610 million IPO in Shanghai, on track to become China's first publicly traded humanoid robotics company. It's targeting 20,000 humanoid shipments in 2026, up from 5,500 the year before. That's not a hobby. That's a manufacturing flywheel spinning up, and soon ordinary investors will be able to own a slice of it on a public exchange.
The factories buying them. UBTech, another Chinese maker, began mass-producing its industrial Walker S2 humanoid in late 2025 and has already booked orders exceeding 800 million yuan (about $112 million). Its robots are working assembly and inspection lines at BYD and NIO, with partnerships spanning Geely, FAW-Volkswagen, Audi FAW, Foxconn, and SF Express. The company is targeting an annual capacity of 5,000 industrial humanoids by 2026, scaling to 10,000 by 2027. These automakers aren't buying robots for the press release — they're buying them because a machine that costs $16,000 and runs three shifts pays for itself against a human salary in well under a year.
The component suppliers. This is the part almost everyone misses, and it's the most durable money of all. Every humanoid needs actuators (the robot's "muscles"), precision motors, gears, sensors, and rare-earth magnets. China controls 63% of the key companies in that global supply chain. So even when Tesla or Figure builds an "American" robot, a large share of the parts — and the profit on those parts — still flows back to Chinese suppliers. Gold rush, picks-and-shovels rules: the people who sold shovels in 1849 did better, on average, than the people who panned for gold.
Why China, and why so fast?
You might reasonably ask: the US has the best AI and the deepest capital markets — how did it lose the cheap robot race so badly? The answer is that a humanoid robot is mostly hardware, and hardware is exactly where three decades of Chinese manufacturing investment pays off.
China didn't win because of cheap labor. It won because it already makes the motors, the magnets, the batteries, and the precision components that every robot needs — and it makes them at roughly one-third of Western cost. Even Jensen Huang, no Beijing cheerleader, called China "formidable" in robotics, noting that its "microelectronics, motors, rare earth and magnets are foundational to robotics" and "the world's best." When the inputs are three times cheaper, the finished robot is too. That's not a temporary discount that a tariff or a subsidy erases next quarter. It's a structural head start.
There's also a flywheel effect that's easy to underestimate. China's electric-vehicle giants — BYD chief among them — are pivoting their enormous mass-manufacturing muscle into humanoids. The same factories and supplier networks that made China the world's EV leader are now being pointed at robots. Morgan Stanley notes the cost of materials to build robots in China dropped 16% in a single year. The leader isn't just ahead; it's getting cheaper faster than the challengers can even start.
Why "robots" belongs on a site about making money with AI
You might be thinking: this is a site about AI, why are we suddenly talking about machines with arms? Fair question. Here's the connection, and it's not a stretch.
A robot is just AI that grew a body. The breakthrough that made the $16,000 G1 possible wasn't the legs or the motors — China has built motors for decades. It was the AI brain: the same kind of large language and vision models that power ChatGPT, now wired into hardware so a machine can see a bin, decide what to grab, and grab it. Huang calls it "physical AI." It's the exact technology this site covers, except that instead of writing your emails, it's stacking your boxes and inspecting your welds.
And the money mechanics are even cleaner than software. With a SaaS tool, you're selling time savings that are notoriously hard to measure, which is why so many AI products struggle to prove ROI. With a robot, the pitch is brutally simple: this machine costs $16,000, a warehouse worker costs $45,000 a year, do the math. That clarity — a physical asset with a payback period you can write on a napkin — is why capital is pouring in faster than into almost any software category.
What this actually means for you
So where do you fit? You're probably not going to out-manufacture Unitree, and you don't need to. The robot boom is creating the same kind of opening the early app stores created — a widening gap between cheap, available hardware and the millions of small businesses that have no idea how to use it.
That gap is the opportunity, and it has three honest shapes:
Be the person who deploys them. Someone has to import these robots, set them up, integrate them into a local business's workflow, maintain them, and explain to a skeptical factory owner why a $16,000 machine beats a $45,000 hire. Right now, almost nobody is doing that job in the West. The people who learn it early — while the technology is intimidating and the supply is Chinese and confusing — will own client relationships that compound for years, exactly like the developers who learned to build apps in 2009.
Rent the hours, don't just sell the box. A robot is an asset that earns a daily wage. Buy a cheap one, rent its working hours to events, warehouses, or retail openings, and you've turned a one-time purchase into recurring cash flow.
Invest in the layer, not the logo. If hands-on work isn't your thing, you can bet on the trend without picking a winning robot brand — through the component and supply layer every brand depends on. Unitree's looming $610M Shanghai IPO is one of the first liquid ways to touch it directly.
China built the cheapest robots in history. The money in your country is in being the person who knows how to land them, deploy them, and explain them to everyone who's still watching the videos and wondering if it's real. It's real. The only question is whether you move while it's still early and awkward — or after the margins are gone.
---
Keep Reading
- Unitree vs Tesla Optimus vs Figure: What a Humanoid Actually Costs in 2026 — the full price breakdown, and why China wins on cost
- The Robot That Replaces a $45K Warehouse Worker for $12K — the ROI math that's driving every purchase order
- How to Make Money Importing Chinese Robots Before Everyone Else — the solo playbook for the robot gold rush
- Why Shenzhen Builds Humanoids 10× Cheaper Than America — the supply-chain edge nobody can copy quickly
