Wang Xingxing, the 36-year-old founder of Unitree Robotics, entered the Hurun China Rich List for the first time on 27 October with a personal fortune of RMB 69.5 billion, ranked 78th. Hurun’s chief researcher called him the year’s biggest story in robotics and the second-largest new entrant on the list.
That number is now being tested by something harder than a ranking. Unitree reported its first set of quarterly results since listing on the Shanghai Stock Exchange’s STAR Market in August, and the market has started to price the company as a business rather than a category.
A share price that already ran the full cycle
Unitree priced its IPO at RMB 150.8 per share. It opened on debut at RMB 1,100, slid to a low of RMB 404.03 in mid-October, then recovered to RMB 739.03 on 28 October. Two days later it closed at RMB 647.05, giving the company a market value of RMB 261.7 billion. The stock fell 4.82 per cent on 29 October and another 2.32 per cent the following day, after the results landed.
Nine months of growth, one quarter of margin squeeze
For the first nine months of 2026, Unitree booked revenue of RMB 1.817 billion, up 55.71 per cent, and net profit attributable to shareholders of RMB 387 million, up 270.54 per cent. Adjusted profit, which strips out non-recurring items, fell almost 20 per cent to RMB 349 million.
The headline profit jump is largely a base effect. In June 2025 a holding vehicle called Shanghai Yuyi issued shares at registered capital price and booked RMB 349 million of share-based payment, all of it charged to administrative expenses in that period. That pushed Unitree’s 2025 net profit down to RMB 278 million and left the first half of 2025 with a net loss of RMB 32 million. The comparison base was artificially low.
The single quarter tells a different story. Revenue rose 69.93 per cent to RMB 665 million, but net profit fell 17.37 per cent year on year to RMB 113 million, roughly half the June quarter. Adjusted profit fell 17.42 per cent to RMB 105 million. Revenue grew, profit did not.
R&D explains much of it. Unitree spent RMB 127 million on research in the quarter, up 249.58 per cent, equal to 19.1 per cent of revenue. For the nine months, R&D reached RMB 263 million, up 191.48 per cent, or 14.47 per cent of revenue. Operating cash flow fell 54.69 per cent to RMB 194 million, which the company attributes to higher purchases, payroll, selling expenses and R&D spending.
The humanoid has already overtaken the robot dog
Unitree built its reputation on quadrupeds. In 2024 they still produced close to 60 per cent of core revenue, while humanoids contributed RMB 107 million, or 27.68 per cent. In 2025 the mix inverted: humanoids reached RMB 868 million, or 51.78 per cent of core revenue, against 41.62 per cent for quadrupeds.
AgiBot overtook Unitree on shipments. According to figures AgiBot released to local media, it shipped more than 8,600 humanoids in the first half of 2026, about 35 per cent of the global market and more than 45 per cent of China, a roughly tenfold increase. Smart Analytics Global estimates Unitree shipped around 5,900 units, up 170 per cent, for 31 per cent of the global market. AgiBot’s first-half revenue was close to RMB 2 billion against Unitree’s RMB 1.152 billion, which grew 48.54 per cent with net profit of RMB 274 million.
Unitree still leads in quadrupeds, with 37 per cent of first-half shipments against 16.3 per cent for AgiBot’s Kubo line, per Counterpoint.
Wang has now said the same thing twice
Unitree shipped the G1+ in September and released UnifoLM-WLA-1.0, a world-language-action foundation model whose embodied reasoning base is built on Qwen3-VL-4B. In October, Wang said publicly on two occasions that his biggest mistake over more than a decade was not believing in AI enough.
That is a striking admission from a founder who spent years arguing that embodiment, not intelligence, was the hard part. The results suggest the market has already moved to the same conclusion. The question for Unitree is no longer whether it can build a robot that walks. It is whether it can build a model that works, and whether the margin survives the attempt.
Revenue up 70 per cent, profit down 17 per cent, R&D up 250 per cent. That is what the transition costs, at least for one quarter.
Editor’s note: This is an adapted translation of the original OFweek report. It has been trimmed and restructured for readability for an international business audience. The full original (in Chinese) is at https://robot.ofweek.com/2026-11/ART-898890-12008-30709637.html.