On August 19, during the Qiqiao Festival, Unitree Technology (688836.SH) will officially debut on the STAR Market, becoming the first pure-play humanoid robot integrator listed on the A-share market. The company set its final IPO price at 150.80 yuan per share on August 6, corresponding to a listing market value of approximately 60.99 billion yuan. The total funds raised are expected to be 6.099 billion yuan, with net proceeds of 5.917 billion yuan after deducting issuance costs, reflecting a price-to-earnings ratio of 219.23 times. Compared to the initial fundraising plan of 4.202 billion yuan, the IPO demand has clearly exceeded expectations. After the market opened, Unitree's share price even surged to 1,100 yuan per share, pushing its total market capitalization close to 400 billion yuan.

However, while the capital market is assigning a high valuation, Unitree Technology is also confronting a practical challenge. The company reported 2025 revenue of 1.699 billion yuan and net profit after deducting non-recurring items of 591 million yuan. Yet, in the first quarter of 2026, despite revenue growing 68.49% year-on-year, its deducted non-recurring net profit fell 52.55%. In the first half of 2026, with revenue up 48.54%, the deducted profit still declined by 19.34% year-on-year. This suggests Unitree is shifting from a phase of rapid volume growth into a new stage where revenue increases do not necessarily translate into profit gains.

More critically, the market believes the true commercial value of humanoid robots must move beyond easy scenarios like scientific research and showcases, and enter industrial sites such as automotive, 3C, and logistics. For Unitree, converting its established cost and shipment advantages into productivity that industrial clients are willing to pay for long-term may be a more significant battle than the IPO itself.

The Journey from Billion to Hundred Billion: Capital Wealth Creation Through Unitree's IPO

Given the generally tightened IPO pace, Unitree Technology's listing progress has been notably fast. Its STAR Market application was accepted by the Shanghai Stock Exchange on March 20, 2026, approved at a review meeting on June 1, and registered by the CSRC on July 1. The company utilized the STAR Market's "pre-review" mechanism, becoming the second enterprise to do so after CXMT (688825.SH).

In terms of issuance, Unitree offered 40.4464 million shares, representing 10% of the post-listing total share capital, at 150.80 yuan per share, corresponding to a market value of about 60.993 billion yuan. The expected total fundraising is 6.099 billion yuan, with net proceeds of 5.917 billion yuan. The issuance price implies a static P/E ratio of 219.23 times based on 2025 earnings, significantly higher than the industry average.

Behind this high valuation lies capital accumulated through multiple funding rounds. Unitree was valued at approximately 13.33 million yuan when founded in 2016. After several rounds, including entry by Sequoia Capital China around 2019 and a nearly 1 billion yuan round in 2024, its post-money valuation reached 12.7 billion yuan in the Pre-IPO round in 2025. By the IPO, its valuation had climbed to 61 billion yuan, creating a significant wealth effect. Founder Wang Xingxing holds 23.82% of shares pre-listing and, through special voting rights arrangements, will control 65.31% of voting rights post-listing. At the issuance price, the book value of his direct and indirect holdings has reached tens of billions of yuan. Meanwhile, early institutional investors and employee stock platforms have also seen substantial paper gains.

It's important to note that IPO wealth creation does not equal realized gains for shareholders. For a newly listed company, these figures represent book values based on the issuance price or current market price. Whether they can ultimately be realized depends on future performance, secondary market valuations, and market liquidity. This situation parallels that of CXMT, which also used the STAR Market "pre-review" mechanism. Both companies possess strong industrial scarcity and attracted high market attention during their IPOs. However, their business models differ significantly. CXMT is a capital-intensive semiconductor manufacturer facing industry cyclicality. In contrast, Unitree's 6.1 billion yuan raise supports much lighter capital expenditure compared to a wafer fab. CXMT relies on long-term capital investment to build capacity, technology, and scale advantages, while Unitree depends more on R&D, supply chain integration, and product iteration. The commonality lies in scarce tech assets being repriced by the capital market; the difference is that one relies on capital-intensive manufacturing, the other on the speed of product commercialization.

Revenue Still Growing, But Why Are Profits Under Pressure?

Behind the IPO spotlight, the market is also focused on changes in Unitree's profitability. From 2023 to 2025, revenue grew from 159 million yuan to 1.708 billion yuan, with deducted non-recurring net profit turning from a loss to 591 million yuan. The gross margin on main business improved from 44.22% to 60.13%. During this period, the company seized the window of rapidly growing demand for quadruped and humanoid robots.

But entering 2026, Unitree's growth rate has begun to shift gears. In Q1 2026, revenue reached 423 million yuan, up 68.49% year-on-year, but deducted non-recurring net profit was only 40.25 million yuan, down 52.55%. In H1 2026, with revenue up 48.54%, deducted profit fell 19.34% year-on-year. The reason for this revenue growth without profit growth is not complicated: the company is spending more to "catch up."

On one hand, Unitree has historically excelled in the robot body, joints, motors, and motion control—what the industry calls the "cerebellum." To enable robots to truly enter factories, they need embodied large models, visual perception, task planning, and autonomous decision-making—the "brain." R&D investment has therefore surged, with H1 2026 R&D expenses reaching 136 million yuan, up 1.52 times from 53.87 million yuan in the same period of 2025. On the other hand, as customers shift from research institutions to industrial clients, Unitree's sales model is changing. Previously, transactions were completed upon product shipment. Industrial clients typically require on-site testing, software adaptation, system integration, and ongoing after-sales support, significantly raising customer acquisition and delivery costs. In H1 2026, sales expenses reached 164 million yuan, already exceeding the full-year 2025 total of 141 million yuan.

This means Unitree must now prove to the market whether its current high gross margins can cover its continuously increasing R&D and sales investments. Competitive pressure adds to the challenge. In 2025, Unitree shipped over 5,500 pure bipedal humanoid robots, placing it at the forefront of the industry. However, domestic robot companies are accelerating mass production. Players like Zhiyuan and UBTECH Robotics (09880.HK) are advancing industrial scenarios, while overseas companies like Tesla are also increasing their efforts. Unitree itself has clearly warned in its prospectus about risks of intensifying competition, slowing revenue growth, and downward pressure on product prices. Therefore, the next phase of competition may not just be about which robot can walk or jump, but who can reliably complete tasks at a reasonable cost.

Several Hurdles Remain for Unitree's Industrial Deployment

Compared to the research, education, and display markets, the industrial market has significantly higher barriers. Currently, the biggest advantage of humanoid robots is their "generality." They can utilize existing factory floors, stairs, shelves, and tools without requiring large-scale production line modifications like traditional robotic arms. But industrial clients don't care whether a robot looks human; they care about efficiency, stability, and return on investment. This is where the difference lies between Unitree and the Hong Kong-listed UBTECH.

Unitree's advantages are clear. Its strong in-house R&D of core components allows for relatively lower product prices. Moreover, years of mass-producing quadruped robots have yielded experience in mechanical structure, motion control, and cost control that can be transferred to humanoid robots. For industrial clients still in the exploratory stage, lower equipment costs mean a lower barrier to trial and error. UBTECH's path differs. It has long focused on robot operating systems, vision, navigation, human-machine interaction, and industrial large models, continuously pushing its Walker series into automotive and manufacturing production lines. In other words, UBTECH invested earlier in figuring out "how to do industrial applications," but at the cost of high R&D spending and greater profitability pressure.

Neither path is inherently superior. Unitree is more like "make the robot cheap, produce it, and sell it first," then gradually build industrial capabilities. UBTECH emphasizes "mastering complex scenarios first," then seeking cost reduction through scale. Ultimately, the decisive factor for both companies will be industrial ROI. Currently, for humanoid robots to become mature industrial equipment, several problems must be solved. First, reliability: factory lines are extremely sensitive to downtime. Occasional AI model errors might be acceptable in consumer scenarios, but on the manufacturing floor, the tolerance for error is completely different. Second, dexterous hands: many industrial processes involve more than just moving items; they require insertion, assembly, grasping, and fine manipulation. Third, battery life and cycle time: if robots need frequent battery swaps or are clearly less efficient than human labor, it's difficult to establish a stable economic case. Finally, there's the after-sales and system integration capability of the equipment supplier. Selling the robot is just the beginning; real industrial delivery often includes software, processes, data, and on-site services.

Therefore, Unitree may not need to directly challenge the most complex workstations like automotive final assembly or precision assembly in the short term. A more realistic path could be to enter scenarios such as inspection, material handling, machine tending, and hazardous environment operations, which require flexibility but have relatively controllable precision requirements. For Unitree, the true watershed may not be the listing day itself, but whether it can transform its "low cost plus high shipment" advantage into capabilities that industrial clients will repeatedly purchase. The IPO provides Unitree with 5.917 billion yuan in funds and places the company under the capital market's spotlight. Going forward, the market will not just watch how fast the company's robots can run, but also whether the enterprise can turn R&D investment into products, products into orders, and orders into stable profits. This may be the real test for the "first humanoid robot stock."