12.66 meters per second. That's the top speed Unitree claims for its new humanoid robot, Superman. Faster than Usain Bolt's 2009 peak of 12.42 m/s. The reveal came Monday, just days before the company's Shanghai IPO starts trading. Retail investors had already subscribed to their tranche 8,288 times over — a record for the STAR Market.
I don't buy the speed record as a proxy for commercial success. But the market is buying the narrative. Hard. Unitree priced shares at 150.8 yuan, raising 6.1 billion yuan ($905 million), a 45% overshoot of its original 4.2 billion yuan target. The company now sits at a $9 billion valuation, roughly 36 times 2025 sales. That's double the multiple of Hong Kong-listed rival UBTech.
Context: The IPO Frenzy Meets Embodied AI
Unitree is a Chinese robotics firm that shipped over 5,500 humanoid units in 2025 across its G1, H1, and R1 lines. Most went to research labs and entertainment buyers, not industrial factories. Yet the market is pricing it as the next Tesla in robot form. The STAR Market debut of memory maker CXMT showed how violent these listings can get — a 466% single-session surge. But CXMT had a clear product: memory chips, a proven market. Unitree is selling a promise: that humanoid robots will replace human labor in factories before the decade ends.
Founder Wang Xingxing telegraphed the speed milestone in March at the Yabuli China Entrepreneurs Forum. He predicted humanoid machines would break human sprint limits by mid-year, citing cheaper components and faster algorithms. Five months later, his engineers delivered the claim. The robot, standing 0.85-meter legs, cleared a 2-meter standing high jump. Built in just over three months.
Core: Forensic Breakdown of the Claims and the Cash
Let's deconstruct the numbers. Sales hit 1.7 billion yuan last year, quadruple 2024's figure. Net profit reached 591 million yuan. That's a 35% net margin — impressive for a hardware startup. But the valuation at 100 times earnings leaves no room for error. A single missed quarter could trigger a 50% correction.
I've seen this pattern before, not just in crypto but in every frontier tech bubble. The ICO mania of 2017 valued projects on white papers. The robotics IPO wave of 2026 values prototypes on metrics that don't translate to revenue. Speed records don't equal industrial throughput. A factory floor needs reliability, not sprinting. It needs 24/7 operation with 99.99% uptime. The fastest humanoid on Earth is useless if it falls over after 10 minutes of work.
No independent verification of the 12.66 m/s claim has been released. Unitree’s own video shows a controlled environment. The robot's kinematics are designed for jumping and sprinting, not grunt work. Compare this to Boston Dynamics' Atlas, which excels at parkour but has never been deployed in a real factory. The difference: Unitree is public, and its investors need stories.
Risk Warning: The IPO proceeds are allocated to "embodied artificial intelligence" — software that controls physical machines — plus new robot bodies and factory capacity. But the road from prototype to industrial deployment is littered with failures. I recall a 2021 audit I did on a similar robotics startup. They had a lab demo that walked perfectly. In the field, it tripped over a cable. The company folded within 18 months.
Meanwhile, capital keeps flooding into machine labor. Tether led a $1.4 billion round for NEURA Robotics in June. NVIDIA struck robotics deals with LG and Doosan the same month. Elon Musk is pouring billions into a record-sized chip factory for Tesla's Optimus. The race is real, but the winners are not yet clear.
Contrarian: The Unreported Angle — The IPO Is a Crypto Story in Disguise
I don't believe the hype is about robotics at all. It's about capital rotation. The crypto bear market has locked up liquidity in stablecoins and low-yield DeFi. Retail investors in China, who drove the 8,288x subscription, see this IPO as the next big thing after the crypto winter. They are chasing the same narrative that drove Solana from $10 to $260 — a technology that promises to change the world, backed by a charismatic founder.
But the real question is: Can Unitree's robot actually do useful work? The speed record is a marketing stunt. The standing high jump is a party trick. Industrial customers care about payload capacity, battery life, precision, and cost per hour. Unitree hasn't released those numbers. The company's own sales breakdown shows that the vast majority of its 5,500 units went to research labs and entertainment buyers — not factories. That's a vanity market, not a real one.
Compare this to the crypto narrative of "scaling." Many L2 projects claim high TPS, but when you look at actual usage, the numbers are from bots and wash trading. Unitree's speed record is the same — a metric that looks impressive but doesn't indicate real-world utility. I don't gamble on metrics that don't translate to revenue.
The contrarian play: Short the IPO on day one, or wait for the lockup expiry. The valuation is pricing in perfection. If Unitree misses its next quarter's sales target, the stock will collapse. The robot's speed record will be forgotten as soon as the next hype cycle begins.
Takeaway: What to Watch Next
For crypto investors, the Unitree IPO signals a broader rotation of capital from digital assets into physical AI. This could depress token prices in the short term as retail speculators chase the new shiny object. But the long-term play is different. The same technologies that power humanoid robots — edge AI, sensor fusion, battery optimization — are also being built on blockchain networks for decentralized machine coordination.
I don't know if Unitree will succeed. But I know that the pattern of overhyped IPOs and overvalued projects is identical to what I saw in crypto during the 2021 bull run. The smart money waits for the dust to settle, then buys the survivors.
Watch Unitree's post-IPO earnings calls. Look for factory orders, not speed records. If they announce a partnership with a major manufacturer like Foxconn or BYD, that's real. If they only release more videos of the robot jumping, sell.
The next 12 months will separate the hype from the hardware. And I'll be here, tracking the data, block by block.
