Unitree Robotics humanoid robot on factory floor

While Western humanoid startups are raising billions on software promises, China's Unitree Robotics just filed for a $610 million IPO on the back of actual mass production. The numbers tell a story the hype merchants don't want you to hear.

Unitree Robotics filed its prospectus with the Shanghai Stock Exchange on March 20, 2026, seeking to raise 4.2 billion yuan ($610 million). The filing reveals something remarkable: a humanoid robot company that actually makes money. Unitree reported 674% profit growth and 335% revenue growth, driven not by venture capital hype cycles but by shipping thousands of units to real customers.

The Numbers Don't Lie

According to the IPO filing and industry reports, Unitree produced 4,200 humanoid robots in 2025. That puts them ahead of Tesla's Optimus production targets, which Elon Musk admitted the company failed to meet. AgiBot, another Chinese competitor, produced 5,100 units in the same period. These aren't pilot programs or demo units—these are mass-produced machines leaving factories and entering the field.

Compare this to the Western approach. Figure AI has raised over $1.9 billion and commands a $39 billion valuation. 1X Technologies has opened pre-orders for its $20,000 NEO robot with deliveries promised for 2026. Sunday Robotics just hit a $1.15 billion valuation for a robot that clears tables. The pattern is clear: Western companies are being rewarded for potential, Chinese companies for production.

The Manufacturing Moat

What Unitree's IPO reveals is that humanoid robotics is bifurcating into two distinct races. The West is running a software race—building better AI models, securing compute partnerships, and raising capital on the promise of general-purpose capability. China is running a manufacturing race—and they're winning.

This isn't just about production volume. It's about the entire supply chain ecosystem that enables rapid iteration at scale. When Unitree needs a new actuator or sensor, they have vendors within driving distance. When they need to retool for a design change, they can do it in weeks, not quarters. The Shenzhen ecosystem that built the world's electronics is now building the world's robots.

The IPO filing also highlights Unitree's vertical integration strategy. Unlike Western competitors who rely on third-party components, Unitree designs and manufactures its own motors, controllers, and key subsystems. This control translates to margin—674% profit growth doesn't happen by accident.

The Droid Brief Take

Here's the uncomfortable truth the venture capital hype machine hopes you'll ignore: manufacturing scale matters more than software elegance in the near term. A robot that can do 80% of tasks reliably and costs $30,000 will beat a robot that can do 95% of tasks unreliably and costs $150,000 every single time.

The West's bet is that software superiority will eventually overcome manufacturing disadvantage. Maybe. But "eventually" is doing a lot of heavy lifting when your competitor is already shipping tens of thousands of units and collecting real-world data at scale. The "Physical AI" moat isn't just about better algorithms—it's about having robots in the field, failing, learning, and improving.

Unitree's IPO is a wake-up call. The humanoid race isn't just about who builds the smartest robot—it's about who can build the most robots, consistently, at a price point that makes economic sense. Right now, that's China.

What to Watch

Unitree's IPO pricing and subscription levels will signal whether public markets value manufacturing execution over software promises. If the offering is oversubscribed, expect a wave of Chinese robotics IPOs to follow. Also watch for Western responses—will Figure, 1X, or Tesla accelerate their manufacturing partnerships, or double down on the software differentiation strategy?

The other metric to track: real-world deployment data. Unitree's robots are entering factories, warehouses, and research facilities now. Their failure modes, maintenance requirements, and actual task completion rates will tell us whether China's manufacturing advantage translates to operational superiority—or whether they're just building prettier paperweights faster.