Unitree Robotics is approaching its highly anticipated stock market debut, and the central question for investors revolves around its post-listing valuation. The buzz around its IPO subscription has been intense, with a popular topic on social media suggesting that securing one allotment could yield a profit of 200,000 yuan. This estimate is based on the average first-day gain of 279.22% for new A-share listings in 2026. Using Unitree's issuance market value of 61 billion yuan and an issue price of 150.80 yuan per share, a profit of 200,000 yuan per lot would imply the company's market value reaching 222.8 billion yuan. While many institutions predict Unitree's market value will surpass 100 billion yuan, with some even forecasting 200 billion yuan, our analysis suggests the capital market tends to overestimate both Unitree and the humanoid robot sector as a whole.

So, what can we expect from Unitree's performance after listing, and can it genuinely support a 100-billion-yuan valuation?

Short-Term Pricing Power Rests in Few Hands

On its first trading day, only 7.44% of Unitree's shares will be available for trading, with the remaining 92.56% locked up. This free-float ratio is exceptionally low for hard-tech stocks on the STAR Market. For comparison, companies like Changxin Technology typically have first-day free floats ranging from 8% to 12%, placing Unitree's 7.44% in the lowest tier historically. The smaller the float, the more volatile the price. On day one, pricing power lies entirely with a select few. With 9.707 million shares from retail investors and 20.3807 million shares from offline placements (90% of which is tradable), only 30.0877 million shares can be traded at any time. This means the opening turnover rate and gains will be driven solely by this 7.44% free float.

Additionally, the final online subscription probability for Unitree is approximately 0.018%, meaning only one in about 5,500 allocation numbers wins a lot, marking one of the lowest subscription rates in STAR Market history. In contrast, Changxin Technology's online subscription rate is around 0.47%, making Unitree over twenty times harder to secure. With such a high proportion of locked shares and low online subscription odds, investors seeking to trade Unitree must do so through the secondary market, creating a liquidity premium in the early listing phase.

The secondary market also has a tradition of speculative hype around robotics listings. For instance, Hong Kong-listed humanoid robot company UBTECH ROBOTICS saw its stock hit an intraday high of 328 HKD in March 2024, with a peak market value exceeding 130 billion HKD, a 264% increase from its issue price. Similarly, there were rumors of Zhiyuan Robotics borrowing a shell listing via Shanghai Weixin New Materials on the STAR Market. Following the expected shell listing, Shanghai Weixin's stock surged with multiple consecutive 20% limit-up sessions, skyrocketing over tenfold in the short term, making it a phenomenon-level target in the A-share market in 2025.

Based on liquidity premiums and the market's sentiment-driven enthusiasm for robotics, we believe Unitree's stock performance is likely to mirror that of UBTECH ROBOTICS and Shanghai Weixin, with significant short-term gains that could approach or break through the 100-billion-yuan market value mark. However, after such surges, both UBTECH ROBOTICS and Shanghai Weixin experienced price corrections, as speculative fervor subsided. As Benjamin Graham famously noted, the stock market is a voting machine in the short term but a weighing machine in the long run. The question remains: can Unitree's fundamentals support a 100-billion-yuan valuation?

Currently a Manufacturing Entity

We believe the key to valuing Unitree lies in how the company is defined. Whether it's flipping somersaults, dancing as a "tech mascot," or performing simple repetitive tasks in factories, Unitree at this stage is fundamentally a hardware manufacturing company. In 2025, Unitree's main business revenue reached 1.676 billion yuan, split across three segments: quadruped robots at 698 million yuan (41.62%), humanoid robots at 868 million yuan (51.78%), and robot components at 104 million yuan (6.19%).

While the product forms among robot manufacturers are similar, what truly determines Unitree's growth potential is its application scenarios. In its response to the second round of inquiry letters, Unitree disclosed that from January to September 2025, its humanoid robot revenue was 595 million yuan, with 73.6% coming from scientific research and education clients, including universities and research institutions. Revenue from commercial consumer scenarios accounted for 17.39% of humanoid robot revenue, while industry applications such as corporate guided tours, smart manufacturing, and intelligent inspection contributed 9.01%. In other words, the current revenue from humanoid robots is driven by research and education, shopping mall demonstrations, and stage performances, essentially selling tech-consumer products to research-oriented buyers.

The capital market is unlikely to pay a premium for robots that only serve "toy-like or display functions." Even in traditional manufacturing, Unitree's rapid growth would typically command a higher valuation. However, we've noticed signs of waning growth. On one hand, the robotics industry harbors a downward pressure from internal competition and product price cuts. According to the prospectus, Unitree's humanoid robot unit price dropped from 593,400 yuan in 2023 to 166,400 yuan in 2025, a reduction of over 400,000 yuan in two years; quadruped robot prices fell from 38,300 yuan to 30,300 yuan. In June 2026, the official price of the R1 humanoid robot was lowered from 39,900 yuan to 29,900 yuan, while competitor Songyan Power's Bumi has already hit 9,998 yuan, and the daily rental price for humanoid robots has plummeted from a peak of 10,000 yuan to under 1,000 yuan.

Price declines are an inevitable part of industry maturation, but they also mean continued pressure on gross margins. Unitree's gross margin fell from 60.7% in March 2025 to 56.01% in June 2026. In terms of profitability, the first half of 2026 saw Unitree's revenue reach 1.152 billion yuan, a 48.54% year-on-year increase, but its non-GAAP net profit attributable to shareholders was 244 million yuan, a 19.34% decline year-on-year. Notably, Unitree's non-GAAP net profit has seen consecutive year-on-year declines in Q1 and Q2 of 2026.

Interestingly, institutions are projecting high growth. According to Wind consensus forecasts, Unitree is expected to achieve revenues of 2.984 billion yuan and 4.532 billion yuan in 2026 and 2027, representing year-on-year growth of 75.62% and 51.88%, respectively. Net profit attributable to shareholders is forecast at 718 million yuan and 1.143 billion yuan, with year-on-year growth of 158.09% and 59.19%. If we calculate based on a 100-billion-yuan market value and 2027 earnings of 1.143 billion yuan, Unitree's price-to-earnings ratio would be 87.48 times, far exceeding traditional manufacturing levels.

Therefore, we argue that Unitree's high growth is uncertain, and viewing the company through a manufacturing lens suggests its 100-billion-yuan valuation carries significant froth. Yet, why is the capital market willing to assign a 100-billion or even 200-billion-yuan valuation?

The Embodied Foundation Model Narrative

If we break down a robot into three components: the body, the "cerebellum" (motion control), and the "brain," the body and cerebellum can be seen as manufacturing and industrial, while the brain, embodied foundation models, falls outside this category. Thus, the capital market's willingness to grant Unitree a higher premium stems not only from its strengths in the body and cerebellum but also from the narrative of embodied foundation models.

At this stage, the core challenge for embodied foundation models lies in data. Similar to the early days of autonomous driving, once the robot industry enters mass production, the primary bottleneck shifts from hardware costs to data supply. Can Unitree's shipment volumes translate into a data moat? As of early 2026, the total globally available real physical interaction data is only about 500,000 hours, less than one twenty-thousandth of the data used for large language model training. Training a general-purpose embodied foundation model requires at least 10 million hours. While the industry often uses simulation and human video to augment data and bridge the virtual-real gap, real-world machines remain the foundation, as robot generalization must ultimately be validated in the physical world.

Embodied foundation models require not just data volume but also data quality, specifically real-machine data, which is closely tied to shipment volumes and scenarios. Players who move first with stronger data feedback loops iterate faster, widening their lead through economies of scale. Currently, Unitree holds a commanding lead in shipments, with over 5,500 humanoid robots shipped in 2025 (pure humanoid, excluding wheeled dual-arm robots), ranking first globally. Unitree's data collection relies on its externally sold robot bodies. Universities, commercial clients, and overseas developers provide real-world scenario collaborations, enabling rapid data feedback. In an open environment, the iteration speed of modules like world models is several times faster than traditional closed-loop R&D models.

However, Unitree's weaknesses are also apparent: it lacks factory scenarios, and its operational depth is not as robust as automotive production lines. We believe that Unitree's ability to pitch a general embodied foundation model to the capital market fundamentally rests on its shipment volumes. If Unitree's robot shipments continue to climb and maintain industry leadership, it has the potential to be the first to train a viable embodied foundation model. This is the only reason the capital market can currently justify a high valuation for Unitree.

Yet, this expectation for an embodied foundation model must materialize within two years to support a 100-billion-yuan valuation, and that appears challenging at present. In summary, we believe that in the short term, Unitree's valuation may approach or exceed 100 billion yuan, but as market sentiment cools, it will likely retreat below that level. A higher valuation would require the successful development of the foundation model.