Staff Reporter |
Editor | Wen Shuqi
China’s embodied AI revolution traces its spark to the explosive rise of Unitree Robotics, and right now, we’re witnessing a milestone that’s genuinely rewriting the industry’s playbook.
On August 19, Unitree Robotics officially rang the bell on the Shanghai Stock Exchange’s STAR Market, pricing its IPO at 150.8 yuan per share. The stock surged an eye-popping 629.44% at the opening bell to hit 1,100 yuan, and was still trading around 900 yuan per share as of this writing — putting the company’s market cap at a staggering 364 billion yuan.
A decade ago, when this company first started hunting for capital, the private markets were still riding the tailwinds of the internet venture boom, and quadruped robots were nothing more than pricey toys for a handful of hardcore geeks. Compared with other investment targets with clearer prospects, Unitree embodied a different kind of allure — the raw, romantic risk of technological idealism.
It wasn’t a glamorous start by any means. Wang Xingxing managed to secure a 2 million yuan angel round from individual investor Yin Fangjun at a valuation of just over 10 million yuan — a decent enough foot in the door. But his “grassroots” background made the early fundraising grind brutally tough. Fortunately, before the industry caught fire in 2025, he’d managed to rally a crew of investors who appreciated his style, thanks to his sheer technical chops and product excellence.
Investors never shy away from praising the rare qualities they see in Wang Xingxing: technically sharp, pragmatic, with an almost obsessive grit for solving problems; commercially savvy, insightful, and relentlessly execution-focused. For many of them, the company’s success feels almost like an inevitability.
But both Wang and his backers might owe a debt of gratitude to AI 2.0 for hitting the reset button on the era. It laid the groundwork, provided the buffer, and ran the simulations needed for Unitree’s meteoric rise and for embodied AI to explode in China — the technological understanding, the commercial imagination, and the investment rhythm all clicked into place. Nobody could have predicted this kind of industry-wide catalytic effect.
From a fringe player selling just dozens of units a year to the global benchmark for the robotics industry, Unitree Robotics is the textbook case every investor dreams of landing in their career: they spotted the right person early enough, and that person rode the wave of the times.
Betting on Technological “Non-Consensus”
Today, humanoid robots and embodied AI are the hottest super-tracks in private markets, but back in 2016, the story was completely different.
Wang Xingxing’s resume lacked the glitter of top-tier universities or star big-tech companies. In his early fundraising days, he hit walls everywhere he turned — a market conditioned to look for elite credentials slammed the door in his face more times than he could count.
In the winter of 2017, Chuxin Capital partner Tian Jiangchuan met Wang Xingxing for the first time at a coffee shop. He showed up with a 40-to-50-page business plan that had absolutely no focus. After that meeting, she jotted down four words in her internal memo: “grassroots background.”
Even though Unitree was already showing its cost-reduction approach on the tech side, Chuxin Capital passed. The core reason? They felt Wang’s background didn’t match what they believed the industry needed, nor their fund’s preferences. It wasn’t until 2020, when they finally grasped the underlying value of the technology, that Chuxin came back and invested at a valuation more than four times higher. Tian later chalked up her initial misjudgment to “elitist arrogance.”
Another investor told this publication that before the industry reached any consensus, he’d connected Unitree with nearly twenty institutions — and not a single one bit.
In a way, that’s the market rewarding those who can break free from inertia.
The early investors shared a common trait: they refused to measure Wang Xingxing against some standardized template. There was no formula to calculate his commercial worth. They simply trusted the man, and the technology he had in his hands.
In July 2019, Wang Xingxing had no idea he was already gaining a reputation in the geek community. Li Yannan — then an analyst, now a managing director at Sequoia China — heard about Unitree through a Zhejiang University senior. He found Wang’s QQ email on the Unitree website, tracked down his WeChat, and reached out.
After that meeting, Li was convinced: Unitree had impressive products, was ahead of the industry in commercialization, and the founder possessed intense technical passion and a hunger to learn. He quickly brought the company to the partners, pushed it through the investment committee, and secured Sequoia China’s seed fund investment in Unitree’s Pre-A round — followed by three more rounds of backing in Pre-A+, Series B, and Series C.
That investment became one of Sequoia China’s textbook cases of full-cycle investing, from seed to late stage. As one of Unitree’s earliest institutional investors, Sequoia China now holds a 7.11% stake, making it the largest independent private equity fund among the company’s shareholders.
Zheng Juncong, founding managing partner at Vertex Ventures, still remembers the 2020 investment committee meeting for Unitree’s Pre-A+ round as “extremely challenging.” He recalls: “At the time, we were thinking — how many units can they actually sell? How big is the market? Everyone figured this company would top out at a 3 billion or 5 billion valuation, max.”
Back then, Unitree was selling only a few dozen units a year, mostly to universities and research institutions. The most direct skepticism from committee members: what commercial potential does this “cute little robot dog that bows” really have?
What ultimately won Zheng over was Unitree’s technological moat — it could hold its own against Boston Dynamics while enjoying a massive cost advantage. And then there was the founder himself, Wang Xingxing.
Back in 2013, while pursuing his master’s degree at Shanghai University, Wang Xingxing deliberately steered clear of the Boston Dynamics-style hydraulic systems that dominated the global high-performance quadruped space. Hydraulic systems packed a punch, sure, but they were expensive, bulky, and a nightmare to maintain. Instead, Wang chose lower-cost outrunner brushless motors and independently built XDog, a fully electric quadruped prototype.
“From an industry perspective, we immediately recognized that this electric drive technology was vastly superior to the mainstream hydraulic approach,” said Huang Jinping, chairman of Rongyi Investment, who previously spent 15 years in wireless communications R&D at ZTE.
For him, it was simple arithmetic: if a robot costs 1 million yuan, there’s effectively no market. But if electric drive can push the price down to a few tens of thousands of yuan, demand and use cases that never existed start to materialize.
After evaluating dozens of robotics companies, Rongyi Investment began participating in Unitree’s funding rounds in 2021, eventually investing a cumulative 66.6 million yuan across three rounds.
This low-cost, highly flexible electric drive approach proved unexpectedly prescient after the AI large-model explosion in 2023. When Silicon Valley’s embodied AI “brains” went hunting for hardware bodies, Unitree’s electric, cost-effective robot dogs turned out to be just about the best option available.
Surviving the Commercialization Gauntlet
Before the grand narrative of general-purpose embodied AI could take root, Unitree first had to solve the very real problem of staying alive — and that, in turn, tested the patience and conviction of its investors.
Huang Jinping describes Unitree’s survival strategy as “laying eggs along the way”: don’t wait around for the perfect general-purpose robot era; sell whatever product you have, whenever you can.
From early applications in power grid inspection, university research, and entertainment displays, to eventually cracking the consumer market, Unitree found commercial footing across a wide range of niche scenarios. Along the way, the technology matured, and cash kept flowing back in.
But productizing and scaling the technology hit the usual wall that plagues Chinese hardware startups: cold shoulders from the supply chain.
“When the company was small, suppliers wouldn’t even consider making custom parts. You’re too small a customer, so they just say no,” Huang told this publication. Unitree’s current self-research rate of over 90% for core components — including joints and LiDAR sensors — was, to a large degree, forced by an unforgiving supply chain.
With no suppliers willing to cooperate, Wang’s team just gritted their teeth and built everything themselves. This reluctant long-term investment ended up building Unitree a formidable moat in both cost and efficiency. Not only did they dramatically slash the cost of joint motors, but they now even sell their self-developed LiDAR units to outside customers.
This ties directly into the company’s capacity for rapid iteration. Li Yannan told this publication: “From day one, Unitree has been rolling out new product breakthroughs roughly every six months, with continuous underlying technological innovation. That’s what drove the evolution from robot dogs to humanoid robots. And the team has insane execution speed — they do everything fast, and that velocity fuels their high-frequency iteration.”
While most companies in the embodied AI space are bleeding money, Unitree flipped to profitability in 2024 — a critical financial milestone that paved the way for its public listing.
This vindicated Wang Xingxing’s early conviction about cost reduction through design. He’s always maintained that design comes first, and mass production second. Many people think scaling up automatically brings costs down — he calls that pure illusion.
At its core, the market’s recognition of Unitree’s technology and products is essentially the market validating the investment decisions of its backers. And as the company’s value grows, some of those backers are naturally looking to lock in their gains.
Vertex kept doubling down in subsequent rounds. “Unitree’s results every single year exceeded what we as investors could have imagined. We rarely see companies this confident, this strong, this impressive,” Zheng said. “By the second and third rounds, we were saying, put in as much as we possibly can.”
Even investors can’t deny that a major factor in Unitree’s leap across the scaling threshold was the “BUFF” it received from the AI era.
Zheng admits he couldn’t have predicted the script that followed. “After all these years in investing, you learn that predicting the big picture is nearly impossible.” The word he keeps coming back to is “inflection point”: “Unitree’s inflection point came with the arrival of AI, and then embodied AI emerged. That’s something I absolutely never saw coming back when I sat on that investment committee.”
Synthesizing information from multiple sources, following the industry explosion, Unitree’s humanoid robot shipments in 2025 grew nearly tenfold year-over-year, while quadruped robot shipments tripled or quadrupled. The industry landscape has been fundamentally transformed.
The Composure of a “Star Company”
Unitree has appeared on the Spring Festival Gala three times, but it was the crucial Year of the Snake appearance that delivered its true “breakout” moment — for both the company and embodied AI as a whole.
At that point, Unitree shares in the private market became nearly impossible to get. One hard-tech investor posted on social media: “Can’t grab any Unitree secondary shares, and can’t finish reading all the robot business plans.” With a minimum ticket of 5 million yuan, investors were practically elbowing each other out of the way. “If you can get in, nothing else matters,” Liu Rong, a partner at Liweiyuan Capital, said in an earlier interview.
The entire sector’s fundraising shifted into overdrive. According to IT Juzi, there were 56 financing events in China’s humanoid robot space in 2024, totaling over 5 billion yuan. A report from the China Academy of Information and Communications Technology (CAICT) and Tsinghua University shows that in 2025, domestic embodied AI and robotics saw 744 investment events, with total funding reaching 73.5 billion yuan.
The heat hasn’t fully cooled in 2026 either. IT Juzi data shows that in Q1 alone, the domestic embodied AI track disclosed over 50 financing events, with cumulative funding of approximately 20 billion yuan — up nearly 60% from the same period last year.
Today, China has more than 20 unicorns in the embodied AI space valued above 10 billion yuan, and five companies with valuations of 20 billion yuan or more. National mega-funds, local government capital, new energy vehicle makers, internet giants, and industry leaders have all piled in.
As a national-level investment fund in the cyberspace sector, China CyberSpace Investment (CNS) participated in Unitree’s early financing in 2022, making it the first national-level fund to back the company. It followed up with two additional rounds in 2024 and 2025.
Behind the glory of being the industry’s star, every wave of Unitree’s viral success has brought multiplied pressure. Fortunately, the company and its founder seem to have kept their heads remarkably clear throughout.
Wang Xingxing has mentioned that the pressure after the Snake Year Gala was immense. “After all these years, I’ve concluded that our only real opponent is ourselves. We need to keep iterating on products and technology at a pace that others basically can’t catch up to.”
At the 2025 Summer Davos in Tianjin, he reiterated that intense competition is a good thing in the long run. “Unitree can withstand the test right now, but we have to keep improving. If our technology stalls, we’ll quickly become a mediocre company.”
This vision of continuous growth, breakthrough, and leadership doesn’t just come from Wang Xingxing himself — in fact, after going public, the entire market is expecting even more from this company.
“Unitree’s moat is shifting from the hardware layer to the ecosystem layer,” Zheng Juncong points out. Just as NVIDIA built an unshakeable software ecosystem with CUDA, Unitree is providing tools and platforms that let developers around the world train algorithms on its hardware.
He compares Unitree to Apple in the embodied AI era: “When the PC came along, everyone could finally develop software on it. We’re now at that same stage of hardware platform establishment.”
Whether this grand vision ultimately comes to fruition depends on how young tech founder Wang Xingxing continues to evolve as an entrepreneur.
Which brings us back to the core of who Wang Xingxing is. Huang Jinping noted to this publication that even when surrounded by the glare of media and capital, Wang maintains remarkable focus on the business and the technology — personally diving into R&D on many critical components. Li Yannan, when discussing the company’s culture, points out that Wang’s current core executive team is the same group from when he first founded the company — not a single person has left.
That might just be Wang Xingxing’s leadership capital as he steers Unitree into its next cycle of challenges.
At a moment when valuation bubbles in the embodied AI track are swelling and deflating, Unitree’s listing has a strangely calming effect on the feverish market. Investors and entrepreneurs each get a lens through which to examine themselves — the former to question whether their decisions are rational, the latter to gauge how much ammunition they can still secure.
Regardless, Unitree’s successful listing has given its early believers a chance to cash in on their conviction. This is the moment when the winners of the venture capital game get paid.
For some early investors who have been in the trenches for years, today’s bell-ringing is clearly not the finish line. Zheng Juncong says current sales volumes don’t even represent 0.1% of the future trillion-yuan market. “The IPO is just the first step. We’re in no rush to exit. We want to keep watching for a few more years.”
