Unitree goes public at a moment when humanoid robotics is moving from staged demos to serious industrial and capital market scrutiny. The Hangzhou based robotics company has filed for an initial public offering on the Shanghai Stock Exchange STAR Market, aiming to raise about 4.2 billion yuan. That makes this more than a company funding event. It is also a test of how public investors value humanoid robotics as a business, not just as a story.
For years, the sector has been driven by spectacle. Robots dancing on television, walking at trade shows, and making viral appearances online helped build awareness. What matters now is whether humanoid robot makers can scale manufacturing, lower costs, improve margins, and convert curiosity into repeatable revenue. Unitree’s filing matters because it offers one of the clearest data points yet on those questions.
The company is already well known in robotics circles for its quadruped robots and has rapidly expanded into humanoids. According to the figures cited around its IPO filing, Unitree became the world’s largest humanoid robot seller last year by volume. That alone does not guarantee a durable lead, but it does make the company one of the most important benchmarks in embodied AI today.
Why Unitree’s IPO matters
This IPO matters for three reasons. First, it gives public markets a rare pure play reference point for humanoid robotics. Second, it shows whether investors believe current demand is strong enough to support high valuations in a category still searching for mainstream commercial use. Third, it reveals how a robotics company with real hardware complexity is trying to build a scalable business model.
Humanoid robotics has attracted intense interest from technology companies, manufacturers, venture investors, and governments. Yet the central commercial question remains open. Can these robots become a broad product category, or will they remain mostly confined to research labs, pilot programs, and tightly defined industrial settings for the next several years?
Unitree’s numbers do not settle that debate, but they move it forward. The filing suggests that demand is no longer purely conceptual. There is real volume, rapidly rising revenue, and improving profitability. At the same time, much of the current humanoid demand still comes from research and education rather than large scale deployment in everyday business operations.
From quadrupeds to humanoids
Founded in 2016 by Wang Xingxing, Unitree first built its reputation with quadruped robots. Those dog like machines helped the company establish strengths in locomotion, control systems, motors, actuators, and integrated hardware design. That base matters because the move into humanoids is not a complete reinvention. It is an expansion built on years of work in embodied movement and robotic systems integration.
Its product portfolio now spans quadrupeds, humanoids, and core components such as motors, reducers, controllers, and sensors. This breadth is important for understanding the business. Unitree is not only selling finished robots. It is also developing the underlying hardware stack that supports performance and cost control.
Between 2022 and September 2025, the company reportedly shipped more than 30,000 quadruped robots. In 2025 it sold around 5,500 humanoid robots. That humanoid figure is the one attracting most of the attention, but the quadruped business still matters because it gave Unitree manufacturing experience, product visibility, and a revenue base before humanoids became its main growth engine.
The numbers behind the IPO
What stands out most in Unitree’s reported financial trajectory is the combination of growth and profitability. Revenue rose from 392 million yuan in 2024 to 1.71 billion yuan in 2025. Adjusted net profit reached 600 million yuan in 2025, a sharp increase from the previous year and an indication that Unitree crossed an important threshold that many robotics companies have not.
This is one of the most notable parts of the story. In robotics, revenue growth alone is common in early phases. Profits are much harder. Hardware businesses must manage manufacturing costs, inventory, supply chains, and after sales support, all while spending heavily on research and development. A robotics company that can show scale and margin improvement at the same time gets closer to being treated as a serious industrial technology business rather than a speculative future bet.
Gross margin reportedly approached 60% in 2025, improving significantly from earlier years. That level suggests that Unitree’s pricing, product mix, and cost structure are all moving in a favorable direction. Humanoids became the largest contributor to core revenue in 2025, accounting for just over half of the total. In 2023, they represented only a very small share. That is a dramatic product mix shift in a short period.
Why prices fell while margins improved
One of the more interesting signals in the filing is that the average selling price of Unitree’s humanoid robots fell sharply, from roughly 593,400 yuan in 2023 to around 167,600 yuan in 2025. Normally, such a drop would raise concern about pressure on margins. Instead, the company’s margin profile improved.
That suggests a few things. Unitree appears to be benefiting from manufacturing learning effects, better component integration, and growing scale. It also points to a strategic choice. Lowering prices can expand the addressable market, especially in education, research, developer ecosystems, and early enterprise testing. In emerging hardware categories, adoption often accelerates when product costs fall enough for more institutions to experiment.
Unitree has said that self development and in house manufacturing of core components give it a cost advantage. If that is sustained, it could become one of the company’s strongest competitive moats. In robotics, the ability to own critical hardware layers matters because external dependency can raise costs, slow iteration, and reduce control over product performance.
Research demand still dominates
Despite the growth in humanoid sales, the current market is still early. More than 70% of Unitree’s humanoid sales last year reportedly went to research and education customers. That is an important reality check.
The market is growing, but it is not yet a mass commercial deployment market. Universities, labs, and developers often buy first in emerging robotics segments because they need platforms for testing, software development, and experimentation. Those buyers are valuable because they help build ecosystems, generate applications, and create feedback loops. However, they are not the same as broad based enterprise adoption.
This means investors should read Unitree’s growth in two ways. On one hand, demand is clearly real. On the other hand, the final shape of sustained end market demand remains unsettled. The next phase is not simply about selling more robots. It is about proving where humanoids deliver recurring value in logistics, inspection, manufacturing assistance, dangerous environments, service workflows, and other operational settings.
A manufacturing story as much as an AI story
Humanoid robotics is often discussed as an artificial intelligence story, but Unitree’s IPO underlines that it is equally a manufacturing story. The company plans to use IPO proceeds not only for research and development, but also for expanding production capacity and building smart manufacturing capabilities.
That focus makes sense. If humanoid robotics becomes a meaningful market, production scale will be as important as algorithmic sophistication. Companies need to build robots consistently, with tight quality control, acceptable lead times, and predictable costs. The winners may not be those with the most attention grabbing demo, but those that can reliably manufacture and improve machines at volume.
Unitree’s vertical integration strategy is central here. Reports around the filing indicate a very high domestic sourcing rate for key components, with more than 90% sourced in China, while some imported materials still make up about one fifth of the broader supply chain. This creates both strength and risk. The strength is tighter control, faster iteration, and lower cost. The risk is exposure to trade policy shifts and dependency on imported high end computing systems, especially where advanced chips are involved.
The China factor in humanoid robotics
Unitree’s listing also says something broader about China’s position in robotics. The country already has deep manufacturing capacity, strong supply chain density, and a policy environment that supports strategic technology sectors. Humanoid robotics fits naturally into that framework because it combines advanced manufacturing, AI, sensors, batteries, semiconductors, and software.
There are now more than 100 humanoid robot companies in China by some estimates. That number is unlikely to hold. As in many fast growing technology categories, the sector will probably consolidate. Public listings can accelerate that process because they create clearer valuation benchmarks, more pressure for execution, and stronger differentiation between companies with real scale and those with mostly narrative momentum.
Unitree may be one of the first companies to define what public market quality looks like in this segment. That does not mean it will dominate the field, but it does mean its performance after listing will be watched closely by founders, investors, suppliers, and competitors.
The competitive landscape is getting tougher
Unitree is not entering a quiet market. Globally, the humanoid field includes Tesla, Figure AI, Agility Robotics, Boston Dynamics, and a growing number of Chinese contenders such as UBTech, Fourier Intelligence, AgiBot, and others. Some are focused on general purpose humanoids. Others are narrowing in on specific environments or task categories.
That matters because the next stage of competition will not be decided only by technical capability. It will also depend on software ecosystems, developer tools, safety, reliability, support infrastructure, and integration into real workflows. A robot that can walk smoothly on stage does not automatically become a practical labor platform.
Unitree’s advantage so far appears to come from cost control, manufacturing efficiency, and speed to market. If it can pair that with stronger software capabilities and clearer enterprise use cases, it could become one of the sector’s most influential players. If not, price competition alone may become difficult as more companies enter the field and larger players push their own platforms.
Valuation and investor expectations
Unitree is reportedly targeting an IPO valuation somewhere in the range of 3 billion to 7 billion dollars. That is a wide range, but it reflects the uncertainty of valuing a fast growing robotics company in a category that is still early. Based on 2025 revenue, that would imply a substantial revenue multiple, especially at the upper end.
Whether investors accept that depends on what they believe about three things. First, how large the humanoid market can become over the next five to ten years. Second, whether Unitree can maintain margin strength as volumes grow and prices continue to fall. Third, whether the company can move from research led demand to broader commercial deployment.
Public markets are often less patient than private capital when categories are immature. That means Unitree may face volatility after listing even if the IPO is well received. Strong growth can support a premium, but any sign of slowing orders, weaker margins, or delayed commercialization could quickly change sentiment.
What the filing says about the next phase of robotics
Unitree’s plans over the next five years are ambitious. The company is targeting annual production capacity of 75,000 humanoid robots and 115,000 quadrupeds. Those numbers imply confidence that demand will expand sharply from current levels.
Even if those targets are not reached on schedule, they show where management believes the market is headed. The company is planning for scale, not niche stability. That is a meaningful distinction. It suggests Unitree sees humanoids becoming a recurring product category with enough demand visibility to justify major manufacturing investment.
The stronger signal, however, is not the target itself. It is the fact that a company is willing to put those ambitions into a public market process where investors can evaluate them against actual financial performance.
