Unitree Just Cleared a $619M IPO — And It's Already Profitable

On July 3, China's securities regulator signed off on Unitree Robotics' plan to raise 4.2 billion yuan — about $619 million — on Shanghai's STAR Market. The headline writes itself: a humanoid-robot company is going public, the embodied-AI era has a ticker. But the number that should stop you isn't the size of the raise or the roughly $6.2 billion valuation it implies. It's a line further down the filing: last year, Unitree turned a profit.

A robotics company that makes money

Unitree reported fiscal-2025 revenue of 1.699 billion yuan (~$250 million) and net profit of 278 million yuan (~$41 million). In a sector defined by cash bonfires — Figure, Tesla's Optimus program, Agility, the entire Western humanoid cohort spending billions to reach production — a Chinese company just walked into an IPO already in the black. That single fact reframes the whole race.

Everyone else in humanoids is selling a future. Unitree is selling a business. The Hangzhou company makes humanoids, four-legged robots, and robot components, and it has been shipping the quadrupeds and lower-cost humanoids in real volume — to universities, research labs, industrial buyers, and increasingly to consumers — for years. That installed base is what turns into revenue, and revenue is what turns into the thing almost no one else in this field has: a P&L that works before the humanoid dream fully arrives.

Speed as a policy signal

The timeline is its own headline. Unitree filed on March 20, the STAR Market listing committee cleared it in early June, and the CSRC confirmed on July 3 — roughly 73 days end to end. IPO approvals in China are not fast by default; they are a discretionary tap that opens and closes with Beijing's priorities. A 73-day clearance for a robotics maker is a policy statement rendered as a calendar.

The message: embodied AI is a national-champion category, and the capital markets are being pointed at it deliberately. Where the United States debates humanoids through venture rounds and Tesla earnings calls, China is wiring a domestic funding pipeline straight into the public markets, giving its robotics champions a currency — a liquid stock — to hire, acquire, and scale against. That's a structural advantage that has nothing to do with the quality of any single robot.

Where the money goes

Unitree earmarked the proceeds for four things: robot AI-model research, robot-body (hardware) research, new product development, and a dedicated smart-robot manufacturing base. Read that list as a strategy. The "AI model" and "robot body" split is the central tension of the entire field — brains versus bodies — and Unitree is funding both sides at once, then pouring the rest into manufacturing capacity.

The manufacturing base is the tell. The humanoid war will not be won by whoever demos the most fluid backflip; it will be won by whoever can build reliable machines by the tens of thousands at a price the market will bear. Unitree already competes on cost in a way Western makers structurally cannot, and a purpose-built factory funded by public capital is how it presses that edge. The IPO isn't a trophy — it's tooling money.

The number that unsettles the West

For a year, the Western humanoid story has been about heroic capex and distant unit-per-hour milestones: Figure ramping BotQ, Tesla converting a Fremont line toward a million-unit annual target, everyone racing to prove they can eventually manufacture at scale. Those are real achievements. But they are all denominated in spending — burn today for volume tomorrow.

Unitree's filing quietly inverts the frame. It is smaller than the Western giants in ambition and in valuation, and its robots are not the ones headlining the flashiest labs. Yet it is the one arriving at the public markets profitable, at scale, and cheap — the three attributes that historically decide who owns a hardware category once the novelty burns off. Consumer electronics, drones, solar, EVs: the pattern is the same each time. The company that wins isn't the one with the most impressive prototype. It's the one that figured out how to make the thing profitably, in volume, for less — and usually it's the same country.

There are caveats worth keeping. Unitree's marquee revenue still leans on quadrupeds and lower-end units, not the humanoid flagships that define the frontier narrative, and a STAR Market listing is a largely domestic capital event with limited reach into Western markets. Profitability at this stage says as much about disciplined, hardware-first pricing as it does about humanoid dominance. This is not proof that Unitree has "won" anything.

But the direction is unmistakable. The CSRC just handed a profitable, high-volume, low-cost robotics maker a war chest and a public currency in under three months. The West's humanoid leaders are still spending their way toward the starting line Unitree crossed a while ago. When the embodied-AI market finally clears its hype and settles into the boring, brutal economics of manufacturing at scale, the company that showed up already making money is the one to watch — and this week it rang the opening bell.