A machine just broke the fastest human stride ever recorded. Unitree’s new humanoid, called Superman, hit 12.66 meters per second on Monday. That beats Usain Bolt’s 2009 peak of 12.42 m/s. The reveal lands days before the company’s IPO starts trading on Shanghai’s STAR Market. Investors have already pushed the share sale to a record subscription level—retail buyers covered their tranche 8,288 times over. The deal values Unitree near $9 billion, or about 36 times 2025 sales. For context, rival UBTech trades at roughly 18 times sales in Hong Kong. The price stretches conventional yardsticks, but the narrative is clear: humanoid robots are the next big story, and capital is racing to catch up.
I’ve seen this pattern before. In 2020, DeFi Summer was a narrative explosion—yield farming, liquidity mining, everyone chasing the next pool. The mechanics were simple: incentives drove behavior. Now the same logic applies to embodied AI. The story is about machines replacing labor, but the underlying driver is the same—capital hunting for a story that can sustain multiples. The Unitree IPO is a clean signal: the market is willing to pay 100+ times earnings for a narrative that hasn’t yet proven industrial adoption. Sales hit 1.7 billion yuan last year, up 4x from 2024, and net profit was 591 million yuan. That’s real revenue, but the valuation still demands a future that may not arrive as fast as the robot runs.
Context: The Narrative Cycle of Humanoid Robotics
Humanoid robots have been a “next year” story for a decade. Boston Dynamics showed impressive demos but never turned them into a business. Unitree, founded in 2016, took a different approach: cheaper components, open-source software, and a focus on research labs and entertainment. By 2025, they shipped over 5,500 units across the G1, H1, and R1 lines. Most went to universities and hobbyists, not factories. The IPO proceeds—$905 million, 45% above the original target—are earmarked for embodied AI, new robot bodies, and factory capacity. The company is betting that the speed record translates into industrial orders. But the disconnect between hype and hardware is wide.
During the 2022 Terra collapse, I watched a narrative detach from reality in hours. The algorithmic stability story broke when the mechanics failed. The same could happen here. The speed record is impressive—2-meter standing high jump on 0.85-meter legs, built in three months—but no independent verification has been released. Founder Wang Xingxing predicted the milestone at the Yabuli China Entrepreneurs Forum in March, citing cheaper components and faster algorithms. Five months later, the claim is delivered. But a sprint record doesn’t mean the robot can weld a car door or care for an elderly person. The industrial narrative is still a prototype.
Core: The Narrative Mechanism and Sentiment Analysis
Let me map the incentive-driven causality. The IPO demand is a symptom of a broader narrative: humanoid robots as the next frontier of AI. This narrative is fueled by parallel events: Tether led a $1.4 billion round for NEURA Robotics in June; NVIDIA struck robotics deals with LG and Doosan; Elon Musk is building a record-sized chip factory. Capital is flowing into physical AI, and the Unitree IPO is a liquid proxy for that thesis. The STAR Market’s structure amplifies the sentiment—retail investors can bid up shares with extreme leverage, as seen with CXMT’s 466% single-session surge. The 8,288x subscription ratio is a liquidity event, not a rational valuation signal.
Based on my experience auditing ERC-20 contracts during the 2017 ICO boom, I know that narratives often precede technical feasibility. The DragonCoin contract I audited had a critical integer overflow that would have allowed unlimited token minting. The team patched it, but the narrative had already raised $12 million. The same dynamic is at play here: the speed record is a narrative catalyst, not a business milestone. The real question is whether the robot can generate recurring revenue from industrial clients. Unitree’s current customers are mostly research labs—they buy one or two units, not fleets. The IPO proceeds will be used to build factory capacity, but that takes time. In a bear market, cash flow matters more than vision.
Contrarian: The Sprint Record Is a Distraction
Here’s the counter-intuitive angle: the robot’s speed is irrelevant. Humanoid robots don’t need to outrun Usain Bolt to be useful in a warehouse. They need to walk, balance, and manipulate objects reliably. The focus on athletic performance is a marketing play designed to capture headlines before the IPO. It worked. But the real risk is that the narrative overshoots the technology’s maturity. In 2026, I ran an experiment with an AI agent that negotiated data access fees on Ethereum. The agent was clever, but it failed in edge cases. The difference between a demo and a production system is where most narratives die.
I don’t care about the price until I see the code. For Unitree, the code is the robot’s control software. The company hasn’t open-sourced the Superman’s firmware, so we can’t verify the claims. The lack of independent verification is a red flag, especially when the IPO is priced at 36x sales. Compare that to the crypto market, where token valuations often trade at multiples of revenue (if any). The same pattern emerges: investors buy the story, then the story breaks. The contrarian trade is to bet that the narrative will cool once the IPO trades and the speed record fades from headlines. The next narrative will be about which companies can actually deliver on the promise of embodied AI, not which one runs fastest.
Takeaway: The Next Narrative
Forward-looking judgment: the Unitree IPO will be a bellwether for the humanoid robot narrative. If the stock surges on debut, it will fuel further capital into robotics startups, including those in the crypto space (e.g., decentralized AI agent networks). If it falls, the narrative will shift to “reality check.” The crypto market will see a similar correction in AI agent tokens, which have been trading on hype rather than product. The question is not whether humanoid robots will eventually matter—they will. The question is whether the current narrative is priced for perfection. Based on the 8,288x subscription ratio, I suspect it is. The sprint record is a great headline, but the real race is on the industrial floor. And that race hasn’t started yet.