Why has Unitree Robotics' cost control sparked discussions? KTX breaks down the logic behind low-cost robots
The cost control of Unitree Robotics has recently suddenly become the focus of market discussion.
The reasons go beyond just the stock price fluctuations after the company's listing. More importantly, a recent in-depth report in *Caijing* on Unitree Robotics and its founder Wang Xingxing has brought the internal management practices, product strategies, and cost control methods of this star robotics company to the forefront.
According to the report, some employees mentioned that at Yushu Technology, reimbursement exceeding 100 yuan may require Wang Xingxing's personal approval; R & D personnel are often criticized; Wang Xingxing himself will be deeply involved in product, promotional videos, and even some detailed decision-making. At the same time, many robot hardware engineers believe that Yushu's real outstanding advantage is not only the pursuit of the highest precision, but also the reduction of robot manufacturing costs through scheme selection, structural design, and rapid delivery.
On September 1, in response to trending topics such as "Reimbursements over 100 yuan require Wang Xingxing's approval" and "The reward and punishment mechanism features almost only penalties with few rewards", Unitree Robotics responded:"Many of these claims are untrue, do not take them seriously. " Therefore, the employees' related descriptions are more suitable as creatives for media reports and community discussions, and cannot be directly regarded as the management system officially confirmed by Unitree. What is truly worthy of market research is instead another issue behind these discussions:
Why can Unitree drive down the prices of its robots while maintaining a relatively high gross profit margin?
The real core of Unitree Robotics may lie in "cost reduction starting from the design phase"
If you only look at online discussions, it's easy to interpret Unitree Robotics' cost control as "the company being overly stingy. "
However, from the perspective of the robotics industry, things are not that straightforward.
Multiple robotics hardware engineers interviewed by Caixin believe that Unitree's advantage does not lie in simply purchasing cheaper components, but in avoiding paths that would be difficult to reduce costs in the future as early as the initial product planning and structural design phase.
For example, a key cost driver for humanoid robots is the joint module, which incorporates a large number of components such as motors, reducers, and encoders.
If a solution with extremely high costs is adopted from the very beginning, the cost reduction that can be achieved even with subsequent production scale expansion will be limited.
Unitree's approach is closer to:
First design a solution that can be mass-produced, then continue to reduce costs through supply chain optimization, shipment scale expansion and rapid iteration.
Caixin reports that a CTO of an embodied intelligence company who has disassembled Unitree's products believes that Wang Xingxing does not blindly stack hardware performance. For example, some products do not use high-cost high-precision encoders, but meet product requirements through other solutions.
This model is actually very similar to that of the consumer electronics industry.
It is not necessary for every single component to be the most expensive; instead, the goal is to enable the entire system to achieve sufficiently good performance while keeping the overall BOM cost under control.
This is arguably the truly key to understanding Unitree's "extreme cost reduction" strategy.
Do low-cost strategies deliver tangible results?
Judging from the public financial data, Unitree's cost control has indeed been reflected in its operating figures.
According to the prospectus of Unitree Robotics, the company's operating revenue in 2025 was approximately 1.699 billion RMB, the gross profit margin of Main Business was approximately 60.13%, and the Net Profit was approximately 278 million RMB.
From 2023 to 2025, the company's operating revenue increased from approximately 159 million yuan to around 1.699 billion yuan, representing a compound annual growth rate of 226.78%. As of the end of 2025, the company has 516 employees, including 184 R&D personnel, accounting for about 35.66% of the total workforce.
The original text states that "the entire company has nearly 1,000 employees, " but the data from the official prospectus currently available shows that the total number of Unitree employees stood at 516 as of the end of 2025. Whether the company further expanded to nearly 1,000 employees in 2026 can only be confirmed through new official disclosures, so it is not recommended to include "nearly 1,000 employees" as a confirmed fact in formal articles.
On the other hand, Unitree's gross margin performance is indeed quite outstanding.
According to peer comparison data disclosed in the prospectus, Unitree's overall gross margin in 2025 stood at around 60.44%, higher than the average level of some robot peers listed in the prospectus. The company also cited cost control capabilities as one of its competitive advantages in the prospectus.
In other words:
Unitree's low-cost advantage does not simply mean "selling at low prices". Instead, it strives to reduce robot manufacturing costs as much as possible while maintaining sufficient gross profit margins.
This is also the part that Capital Markets truly care about.
Is "Wang Xingxing controls everything" an advantage or a risk?
One very prominent keyword in the Caijing report is: Wang Xingxing himself.
Reports describe that, in the eyes of some employees, Wang Xingxing is deeply involved in product details, not just algorithms and robot hardware, and sometimes even pays attention to screw lengths, scenarios and product specifics; employees may also need to wait for approval when reporting to him. (Caijing Magazine)
Looking at the early stages of startups, this model where founders personally take charge of product development is not uncommon.
Its advantages are obvious:
The decision-making chain is short, the product direction is unified, and the founder can quickly identify problems and directly make trade-offs between cost, product experience, and delivery speed.
This is especially true for industries like humanoid robots, which are still in a phase of rapid iteration, where hardware solutions can change every few months. A founder who is deeply familiar with the system and directly involved in product decision-making can sometimes boost efficiency.
However, as the company grows larger, this model will also encounter new problems.
If a large number of matters eventually need to be approved by a single person, the founder themselves may gradually become a bottleneck for the organization's efficiency.
This is also why the detail of "people queuing outside the office for approval" in the screenshot of the original text sparked discussions.
The real issue behind it is not "whether approval is really required for 100 yuan", but:
When should a fast-growing tech company transition from founder-led to organization and process-driven?
This issue may be more important than the specific reimbursement amount.
Will extreme cost reduction come at the expense of product quality?
This is also one of the most noteworthy risks associated with Unitree's business model.
Caijing Magazine reported that in certain early stages, Unitree may have prioritized shipment speed and market capture, while planning to resolve product issues through subsequent iterations. Some supply chain sources also noted that Unitree's product design iteration speed is extremely fast. (Sina Finance)
From the perspective of industrial development, this is a very typical trade-off:
It is extremely difficult to achieve the ultimate in performance, cost, quality and delivery speed all at the same time.
If a robotics company requires every single component to meet the highest precision standards in the industry from the very beginning, the final product is likely to be extremely expensive, and the market size will be difficult to expand rapidly.
Conversely, if more emphasis is placed on low cost and fast mass production, it is necessary to compensate for part of the hardware performance gap through structural design, software control and supply chain management.
Unitree's current approach is clearly much closer to the latter.
Judging from the results, this strategy has indeed helped the company rapidly expand its business scale.
However, once humanoid robots enter the stage of truly large-scale application, their reliability, maintenance rate, continuous working time and product lifespan will become increasingly important.
Therefore, what Unitree really needs to prove in the future is not:
"Can robots be made even more affordable? "
Instead:
"Can robots meet the increasingly stringent industrial and commercial application requirements while keeping costs low? "
This is what will determine the competitiveness of Nio in the next phase.
KTX Crypto Watch: What's Really Trading Amid the Pullback in Robot Stocks?
Unitree officially listed on the Sci-Tech Innovation Board on August 19,2026. According to an announcement from the Shanghai Stock Exchange, the company's stock code is 688836.(China Southern Power Grid)
Following its listing, Unitree once secured an extremely high market valuation.
Caijing reported that the company's market capitalization once reached about 440 billion yuan after its listing, but as of August 28, its market capitalization stood at around 248.8 billion yuan, having fallen significantly from its peak.
At the close of trading on September 1, Unitree's share price stood at around 571.1 yuan, with a market capitalization of approximately 231 billion yuan.
For KTX Crypto, what is more noteworthy about this matter is not the specific rise or fall of any individual Class A Share company, but the valuation of AI and robotics is gradually shifting from the "storytelling phase" to the "monetization validation phase".
In the past, the market was willing to assign very high valuations to humanoid robots, with the core expectation being:
Improved AI capabilities → smarter robots → mass production → lower costs → explosion of commercial use cases.
However, once the company is actually listed on the stock market, the market will gradually start to ask:
Is income growing fast enough?
Where are the actual application scenarios of robots?
Are profits sustainable?
Can the cost advantage be sustained?
Has the valuation already priced in the growth for the next few years?
This actually follows a similar logic to the AI, DePIN and robot-related narratives in the Crypto market.
While prices in the early stages may be driven more by expectations, the ultimate determinants of long-term value remain products, users, revenue and real demand.
The KTX current market page provides different market categories such as Crypto, TradFi, and Stocks, which can be used to observe changes in market sentiment across different risk assets.
For users who have not yet used KTX, you can create an account through the official KTX registration page to learn more about products in different markets. The actual listing status of specific assets shall be subject to the real-time page of KTX.
FAQ
Does Unitree Robotics actually require Wang Xingxing's approval for any reimbursement exceeding 100 yuan?
A report by Caijing cited an employee making this claim, but Unitree Robotics responded on September 1 that "much of the content" circulating online is untrue and should not be taken seriously. As such, it cannot currently be regarded as an officially confirmed company policy.
How many employees does Unitree Robotics currently have?
According to the official prospectus, as of the end of 2025, the company had a total of 516 employees, of whom 184 were R&D personnel, accounting for 35.66%. The latest employee count for 2026 is subject to new official disclosures.
Why can Unitree keep robot prices relatively low?
Public reports and prospectuses show that its advantages stem more from product solution selection, structural design, supply chain control, large-scale shipments and rapid iteration, rather than simply purchasing cheaper components.
Has Unitree Robotics been listed?
Yes. According to the announcement from the Shanghai Stock Exchange, Unitree Robotics was listed on the Sci-Tech Innovation Board on August 19, 2026, with the stock code 688836.
Does the drop in Unitree Robotics' stock price signal that the robotics industry is in decline?
Such a conclusion cannot be drawn. The stock price of a single company is also affected by IPO valuation, capital sentiment, earnings expectations and market risk appetite. The long-term development of the robotics industry needs to be observed in terms of actual shipments, application scenarios, revenue growth and technological iteration.
Does low cost necessarily lead to a decline in robot quality?
Not necessarily. Cost reduction can come from structural optimization, scale effect and supply chain efficiency, rather than simply cutting investment in quality. However, excessive pursuit of delivery speed and cost may also increase pressure on reliability and quality control. Therefore, judgment needs to be made in combination with the repair rate, service life and customer feedback.
Conclusion
The truly valuable part of this round of discussions on Unitree Robotics' cost control may not be "whether Wang Xingxing approves a 100-yuan reimbursement or not. "
These employee accounts remain controversial, and Unitree has explicitly responded that many of the claims are untrue.
What is more certain is:
Unitree has indeed developed a very distinct low-cost, high-performance, fast-iteration approach.
In 2025, the company's revenue approached 1.7 billion yuan, the gross profit margin of its Main Business exceeded 60%, and R&D personnel accounted for more than one third. (DataClouds)
This shows that "low cost" itself does not mean a lack of technical content.
On the contrary, in the hardware industry, the real challenge is often not about stacking the most expensive components, but finding a scalable balance point between cost, performance, reliability and delivery.
For Capital Markets, what needs to be verified next is not whether Unitree can still create hype, but whether it can continue to translate its low-cost advantage into scale, profits and real-world application scenarios.
This is the real investment logic behind Unitree Robotics' cost control that deserves long-term attention.
Included Original Text
This article on finance and economics is quite interesting. As the share price of Unitree Robotics, a leading robotics company, continues to fall, various employees have started to share their firsthand experiences.
The company has nearly 1,000 employees, but any reimbursement exceeding 100 yuan must be approved personally by the boss, Wang Xingxing.
An employee revealed the company's reward and punishment mechanism, which only imposes penalties without offering any rewards. Most R&D staff have performance scores below 1, and being criticized is a common occurrence.
All external videos produced by the company must be approved by the boss Wang Xingxing, who often scrutinizes every detail of the scenes, shots, and lighting.
When it comes to compensation, although it is a leading company, the average pay raise is generally lower than the industry average, and the workload is quite demanding.
The company places extreme emphasis on cost control. Instead of pursuing the highest quality and precision, it focuses on extreme cost reduction, and rapidly iterates hardware R&D through shipment scale and delivery speed.
Stocks, the technology sector, and the cryptocurrency market all carry significant volatility risks. Some of the internal management information referenced in this article is sourced from media interviews and community compilations, and related claims may contain incomplete information or disputes. Official company disclosures and regulatory documents shall prevail. This article is for informational and educational purposes only and does not constitute any financial or investment advice.
