Gasgoo Munich-A single new listing has generated significant activity in the STAR Market.
On the evening of August 10, Unitree Robotics announced the final winning rate for its online offering: 0.01809759%. This indicates fewer than two winners out of every 10,000 subscription accounts.
The potential payout is notable. Based on the average first-day gain of 276.04% for A-share listings since 2026, a single lot of Unitree could yield a paper profit exceeding 200,000 yuan. If it matches the STAR Market’s average first-day increase of 466.61% this year, that profit could reach 351,800 yuan.
Social media has featured discussions on "making 200,000 yuan on a single lot" and predictions that the "first humanoid robot stock is ready to rise." This niche, specialized sector in hard technology has suddenly triggered widespread interest in IPO subscriptions.
However, this strong interest in IPO subscriptions signifies more than the performance of a single listing.
As the first pure-play humanoid robot listing on the A-share market, Unitree’s debut establishes a pricing benchmark for the entire embodied AI sector. Humanoid robots are now an industrial segment with a tangible market capitalization and performance anchors, rather than concepts in research reports or displays at trade shows.
As interest increases, the need for clarity grows. Behind this widespread activity lies a critical question: Is this reasonable pricing for the start of industrialization, or an expectation overdraft driven by emotion?
Why Unitree? What is Driving This Interest?
Many attribute Unitree’s attention to its "first humanoid robot stock" designation. That is the result, not the cause.
The A-share market has never lacked "firsts." However, triggering widespread subscription interest requires a combination of three factors: sector scarcity, timing in the industrial cycle, and solid fundamentals.
First is the "scarcity of targets" in the A-share hard technology space.
Over the past two years, the humanoid robot concept has seen periodic attention, yet capital could only flow upstream to harmonic reducers, servo motors, and sensors. There was not a single play focused on complete humanoid robots with scalable shipments.
It mirrors the early days of the new energy vehicle expansion: investors focused on lithium batteries and components until the leading automakers went public, finally providing an anchor for the sector.
Unitree fills that void. It is the first company to scale humanoid robot production to over 5,500 units annually, generate more than 800 million yuan in revenue, and take the leading position in global shipments. It enters with an established position in sector leadership.
Image Source: Unitree Robotics
In a zero-sum A-share market, scarcity is often the strongest logic for a rally. This year, hard technology stocks that offer long-term industrial potential, short-term performance, and technological barriers are scarce. Capital naturally clusters around the few certainties.
Unitree’s emergence provides the entire market with a "standardized instrument" for investing in the humanoid robot sector.
Second, Unitree’s listing timing coincides with a critical window in the industry’s iteration.
The global humanoid robot industry is currently in the "initial phase of commercial pilot verification." It is not a purely conceptual stage with no clear future, nor a mature stage of mass adoption. It sits at the intersection where technology is deployable, mass production is achievable, and scenarios are explorable.
Leading companies at this stage are best positioned to enjoy a dual valuation premium: one based on performance delivery, the other on future potential.
Two years ago, the market debated whether humanoid robots could walk steadily. Today, the consensus has shifted to "humanoid robots will be deployed; it is a matter of when."
Unitree’s entry into the capital market at this juncture captures the market’s collective expectations for the sector’s long-term value.
Third, Unitree has a sufficiently solid fundamental "safety cushion."
Unlike many hard technology firms that debut with losses, Unitree is not relying solely on conceptual narratives.
According to the prospectus, Unitree generated 1.699 billion yuan in revenue in 2025, with net profit excluding non-recurring items approaching 600 million yuan and a gross margin exceeding 60% for its main business. Revenue from quadruped robots has increased in recent years, reaching 120 million yuan in 2023, 230 million in 2024, and 697 million in 2025 — representing year-on-year growth of 93.12% and 202.60%, respectively.
During the same period, revenue from humanoid robots increased from 2.9671 million yuan to 107 million yuan and then to 868 million yuan. Its share of the main business rose from 1.88% in 2023 to 51.78% in 2025.
This means Unitree’s business structure combines a "mature quadruped business providing a floor" with a "high-growth humanoid business opening the ceiling." Even if the humanoid business lags expectations, the stable cash flow from quadrupeds offers fundamental support, making its risk profile lower than pure-concept plays.
This "room to run, floor to stand on" attribute aligns with the market’s current cautious risk appetite.
Finally, the issuance structure amplified the market sentiment.
Unitree’s total public offering was over 40 million shares. After deducting strategic placements and institutional subscriptions, the free float available online was limited. Once the clawback mechanism was triggered, the final online tranche was 9.707 million shares. With a low subscription threshold, retail investors could participate easily, creating a "herd effect" where investors sought access.
Amplified by online labels like "first humanoid stock" and "guaranteed winner," the attention increased.
Roadshow Scrutiny: Is the Valuation an Overdraft?
However, the stronger the subscription sentiment, the easier it is to mask genuine controversies.
Unitree’s recent online roadshow for its STAR Market debut became a focal point for these market divisions.
During nearly eight hours of Q&A, investor questions covered performance growth, technical shortcomings, supply chains, and overseas risks. While seemingly scattered, every sharp query converged on a single core issue: Does Unitree’s issuance valuation match its current development pace? And how will it support such a high valuation in the future?
Image Source: SSE Roadshow Live Screenshot
Analyzing these doubts, the most intuitive and easily verifiable concern is the sustainability of growth momentum.
From 2023 to 2025, Unitree’s revenue increased from 159 million yuan to 1.699 billion yuan — a tenfold increase in two years. That is the core confidence supporting its high valuation.
However, for the first half of 2026, Unitree projects revenue between 1.052 billion yuan and 1.128 billion yuan, representing a year-on-year increase of 35.62% to 45.41%. Net profit is expected to range from 258 million yuan to 306 million yuan, while net profit excluding non-recurring items is projected at 236 million yuan to 283 million yuan — a year-on-year decline of approximately 21.97% to 6.43%.
Wang Xingxing, Unitree’s Chairman, General Manager, and CTO, responded during the roadshow that the slowdown is due to a higher revenue base, cooling industry interest, and intensifying competition. Additionally, to seize opportunities in embodied AI, Unitree has increased investment in R&D, talent, and sales — factors weighing on growth rates.
However, the market’s concern is this: the high growth of the past two years was essentially a concentrated release of demand from universities and research institutions, a market with a distinct ceiling. Once research demand saturates, and industrial and consumer demand fail to compensate, can high growth continue?
This question was raised repeatedly during the roadshow, yet Unitree offered no clear earnings guidance, choosing instead to reiterate the "vast potential of the industry."
That itself is a signal: short-term performance volatility is likely the new normal for the market.
The second layer of controversy concerns when the "reliance on research" will be broken.
Market speculation has long held that over 70% of Unitree’s humanoid robots are sold to universities and research institutes.
While Unitree has not explicitly confirmed that figure, it acknowledges that research, education, and commercial demonstrations are currently the primary application scenarios.
Wang stated that current global demand for humanoid robots comes mainly from research, education, and commercial consumption. Large-scale industrial deployment is distant. In the short to medium term, applications will remain concentrated in scientific research, development, education, cultural performances, intelligent services, and high-risk emergency operations.
This raises a practical issue: research procurement is often a one-time purchase of tools, with limited repeat business and scale. To support billions in revenue, the company needs industrial-scale bulk procurement and mass-market consumer sales.
Globally, the industrial deployment of humanoid robots is in the pilot stage. Scenarios where they can replace human labor stably are scarce.
"Since robots are not yet mature at the hardware or software level, true large-scale application has not been achieved in industrial, commercial, or household settings," said an analyst at Gasgoo Automotive Research Institute. "Whether bipedal or wheeled, humanoid robots have a long path ahead before mass adoption in any scenario."
Specifically, data from Gasgoo Embodied Intelligence shows that roughly 35% of current humanoid robot shipments go to entertainment and performance applications, and 20% to data collection firms and research institutions. Combined with other commercial service scenarios, commercial applications account for nearly 90% of total volume.
Given the high standards of industrial scenarios, the phase where humanoid robots are "paid for by research and performance" is expected to persist.
If the revenue structure remains focused on research and performance, the market ceiling for humanoid robots will be lower than estimated — and the current valuation suggests an overdraft.
Image Source: Unitree Robotics
The third layer of doubt is whether the shortcoming of "strong motor skills, weak cognitive skills" can be addressed.
Unitree’s motion control capabilities are recognized as a strength, with skills like backflips and high-speed running placing it in the industry’s first tier. However, in the long term, the outcome for humanoid robots will be decided by the "brain" — the environmental understanding, autonomous decision-making, and task execution capabilities enabled by embodied large models.
Wang noted that for large-scale commercial use in industrial and household settings, the key breakthroughs needed are in embodied large models and the precision and durability of "dexterous hands." The primary technical challenge remains that embodied models globally are in an early stage with insufficient generalization capabilities.
While Unitree is pursuing both WMA and VLA technical routes and has partnered with DeepSeek, the gap in large model accumulation remains distinct compared to top-tier players like Google DeepMind and Tesla.
Investors pressed repeatedly on the progress and commercialization timeline of these models during the roadshow, but Unitree mostly responded with phrases like "R&D testing stage" or "industry-leading tier," offering few concrete details.
The market’s concern is practical: if its large model capabilities lag, Unitree risks becoming a hardware manufacturer, missing out on the ecosystem premium of the intelligent age. The valuation logic for a hardware company differs from that of an AI platform.
Finally, there is no avoiding the valuation itself.
The market discussion about a "219x P/E ratio" refers to the pre-exclusion figure; based on net profit excluding non-recurring items, the offer price corresponds to a P/E of about 92 times. That is a high multiple for the broader high-end manufacturing sector, implying an assumption that Unitree will maintain a compound growth rate of over 50% for years to come.
However, high growth and high valuation carry risks. If performance meets targets, it is reasonable pricing for the start of an industry. If expectations fall short, the valuation correction will be significant.
From this perspective, the current market capitalization looks like a "sector option" — a bet on the possibility of Unitree becoming the leader of the humanoid robot era, rather than a reflection of realized value.
Amid the Interest, Unitree Must Clear Three Hurdles
The IPO subscription interest will eventually fade. The listing is the true test.
Under the market’s spotlight, Unitree needs to deliver verifiable progress and continuous breakthroughs across key levels of industrial deployment to sustain its current valuation expectations.
Image Source: Unitree Robotics
The first hurdle is commercial deployment: bridging the gap from the stage to the production line.
Performing and working are entirely different. Executing fluid moves at a trade show is easier than working stably on a factory floor for three months.
"Industrial users prioritize load capacity — the maximum weight the end-effector or dexterous hand can bear," noted a Gasgoo analyst. "The second core metric is operational precision. Processes like visual inspection, dispensing, and assembly often require precision down to 0.0x millimeters at a fixed station. The third is response speed and latency. Assembly lines move fast; once a part is in place, there’s no time to wait seconds for a robot to start. Industrial scenarios demand extremely high real-time response."
This means that while Unitree has reduced prices through full-stack self-development, getting factories to buy in bulk requires more than low costs; the robots must be useful and stable.
Moreover, industrial customers calculate the costs pragmatically: robot cost, number of workers replaced per year, failure rates, maintenance costs, and the return on investment cycle. Current humanoid robots have not crossed the ROI threshold. High-end units cost hundreds of thousands of yuan, and entry-level units tens of thousands. When factoring in debugging and maintenance, the economics of replacing human labor do not add up.
The second hurdle is technological iteration: how long can the motion advantage be defended, and when will the brain catch up?
Motion control is Unitree’s first-mover advantage, but hardware barriers are not permanent. In China, technology for core components like joints, motors, and reducers diffuses rapidly. It may take a year or two before many companies’ robots can walk and run steadily. By then, motion capability will be an industry standard, not a unique moat.
The embodied large model and the data loop will create a generational gap. This is the industry’s shared challenge: collecting real-world scenario data is expensive and hard to label with poor generalization, while simulation data differs from reality.
Image Source: Unitree Robotics
Unitree is pursuing an "open ecosystem + open-source model" strategy, aiming to feed algorithm iteration with data from global developers. The direction is sound, but ecosystem building is slow. Whether it will succeed, and the time required, remains unknown.
Beyond that, issues like product lifespan, thermal management, and reliability — problems that surface after mass production — are difficult challenges. A lab prototype is a different difficulty class from a mass-produced product with an annual volume of 100,000 units.
The third hurdle is the competitive landscape: as cross-sector giants enter, is the moat deep enough?
The humanoid robot sector involves more than startups. Overseas, Tesla and Figure are applying AI and supply chain advantages. Domestically, Xiaomi, BYD, and XPENG are entering the market.
These cross-sector players have advantages: automakers possess mature supply chain management, mass manufacturing experience, cost control capabilities, and their own factories for scenario deployment; technology companies bring accumulation in large models and algorithms.
Many say Unitree’s rivals are other startups, but the real threat comes from these giants. Once they replicate their automotive supply chain capabilities in the robotics sector, the cost reductions driven by scale could be significant, potentially triggering a price war. Whether Unitree can maintain its cost-performance advantage is an open question.
Unitree has its own moat: first-mover brand recognition, a large developer ecosystem, years of accumulated motion control data, and a focused team. However, these advantages are not unassailable. Against the resource influx of giants, the window of opportunity is short.
Conclusion
Unitree’s listing is a milestone for the humanoid robot industry.
Its core value lies in establishing a public value benchmark for the entire sector. Future primary market financing, upstream and downstream pricing, and technical route comparisons will reference Unitree. The good and the bad will be amplified by the secondary market, pushing the entire industry toward more rational development.
However, perspective is required: an IPO is a starting point, not a final destination.
The IPO interest will fade, and stock volatility will settle. A company’s value is decided by simple factors: the number of robots working in factories sold next year, when embodied models allow robots to understand the physical world, and whether it can maintain its position and ecosystem amidst competition.
The road for humanoid robots is in the early stages. After the interest subsides, the true race of technology and commercialization begins.
