Capital markets always favor the next grand story. On September 1, 2026, Mech-Mind Robot (HKEX: 09615) listed on the Hong Kong Stock Exchange as the "first listed embodied intelligence eye-brain-hand company," with its IPO oversubscribed 3,835 times, setting a new record for Hong Kong robotics IPOs.
Yet the contrast between the frenzied 3,835x pre-listing oversubscription and the post-listing share price pressure has torn away a layer of the capital narrative. As of the September 28 close, its shares traded at HK$79.5, a discount of over 21% from the IPO price of HK$101.7.
On one side is the capital frenzy of a 3,835x oversubscription; on the other is the persistent post-listing market indifference. Behind this valuation rift lies a question that must be answered: does Mech-Mind Robot's approximately 30x PS at issuance represent a reasonable pricing of the future of embodied intelligence, or an overly optimistic overvaluation? We may be able to glean some clues from its latest financial results, its ability to deliver on technology promises, and its market outlook.
Two Sides of the Earnings Report: Impressive Growth, But Profitability and Scale Effects Yet to Materialize
On September 24, Mech-Mind Robot delivered its first interim report card since listing: revenue, orders, and overseas income all grew strongly, with gross margins continuing to rise — the growth curve is impressive enough. However, peeling back the prosperity of the data, the other side is equally clear: profitability remains elusive, and scale effects are still a check that has yet to be cashed.
In the first half, the company's revenue reached RMB 237 million, up 54.7% year-on-year; new orders totaled RMB 335 million, surging 75.3% year-on-year; overseas revenue was approximately RMB 100 million, up 68.3% year-on-year, with overseas markets having become an important growth engine. During the period, gross margin rose from 61.4% to 65.0%, with the product's own profitability continuing to optimize. Customer quality is equally outstanding, with a repurchase rate of 93.9% in the first half, serving over 100 Fortune Global 500 companies and covering nearly 50 countries and regions.
Beneath the impressive growth curve, profit pressure remains heavy. In the first half, the company's operating loss widened to RMB 77.454 million, with R&D investment of RMB 80.1 million, up 72.9% year-on-year, and R&D expenses accounting for 33.8% of total revenue. Taking a longer view, the company's net losses from 2023 to 2025 were RMB 401 million, RMB 283 million, and RMB 360 million respectively, with cumulative adjusted losses over three years totaling RMB 657 million; in Q1 2026, it continued to lose RMB 57 million. Although the adjusted loss narrowed slightly by 5.9%, revenue grew 54.7% while operating losses simultaneously expanded by 18.8%.
It is worth noting that this set of data creates considerable tension with management's claim of "operating leverage clearly emerging." The core logic of scale effects is that revenue expansion drives down unit fixed costs, causing losses to narrow at an accelerating pace as revenue grows. Mech-Mind Robot's current performance indicates that the scale dividend is far from truly arriving.
Of course, changes in Mech-Mind Robot's inventory also deserve attention: it increased from RMB 63.2 million at the end of 2025 to RMB 89.3 million at the end of H1 2026, a rise of 41.4%. Although lower than the revenue growth rate, considering the company's highly standardized products, inventory buildup may still signal that downstream demand is not as optimistic as the order data suggests.
Sales expenses are another signal. From 2023 to 2025, the company's sales expenditure was RMB 186 million, RMB 162 million, and RMB 168 million respectively, consistently exceeding R&D expenses in the same period (RMB 119 million, RMB 109 million, RMB 113 million). In 2023, the sales expense ratio reached as high as 102.7%, meaning that the entire year's revenue was insufficient to cover sales costs. This raises a key question: is the current high growth driven by natural demand from product value, or by orders purchased through sustained high sales investment? Once sales investment contracts, whether growth can be maintained remains unknown.
From the above, it is not difficult to see that when Mech-Mind Robot's order growth cannot outpace the widening of losses, and when the declaration of "operating leverage emerging" conflicts with the reality of an 18.8% expansion in operating losses, the rift between the company's valuation and its fundamentals will not close on its own.
Technological Fog: 99% of Revenue from a "First Embodied Intelligence Stock" Comes from 3D Cameras
The core of Mech-Mind Robot's external narrative is building an integrated "eye-brain-hand" solution for embodied intelligent robots, but the robot guidance business that currently contributes all revenue is essentially a mature set of 3D cameras and supporting software, serving traditional industrial robots to complete object recognition, grasping, and loading/unloading in structured scenarios. This product line has been iterated for over a decade and belongs to the mature solutions of the industrial vision track — it is not an embodied intelligence system for humanoid robots in open environments.
Breaking down Mech-Mind Robot's revenue structure makes the answer even more intuitive. In 2025, the company's total revenue was RMB 389 million, of which intelligent robot guidance products contributed approximately RMB 361 million, accounting for 93%. Intelligent inspection and measurement products contributed approximately RMB 23.07 million, accounting for 5.9%. The two businesses together contributed about 99% of revenue.
In H1 2026, intelligent robot guidance products generated RMB 216 million (91.14% of revenue), and intelligent inspection and measurement products generated RMB 21.2 million (8.9% of revenue), together contributing over 99%. The two products truly branded with the name "embodied intelligence" — the Mech-GPT multimodal large model and the Mech-Hand dexterous hand — together accounted for less than 1.1% of revenue.
According to the prospectus plan, these two products are expected to begin scaled rollout only by the end of 2026, and to generate "revenue of actual commercial scale" only in the second half of 2027. In other words, a company that listed as the "first embodied intelligence stock" derives 99% of its revenue from an industrial 3D camera that has been sold for a decade. This is likely not the early stage of embodied intelligence commercialization — it is the stage before embodied intelligence commercialization has even begun.
This can be glimpsed from the technological paradox of "breakthrough progress." Management claimed in the earnings report that "embodied brain technology has achieved major breakthroughs" and "validated the correctness and effectiveness of the investment direction." Yet R&D expenses in H1 2026 were RMB 80.1 million, up 72.9% year-on-year, accounting for about 33.8% of revenue. A paradox thus emerges: if the technology had truly achieved a "major breakthrough," R&D investment should be entering a phase of diminishing marginal returns; if R&D expenses are still growing at a high rate, it means the technology is far from converging. With only RMB 80 million in R&D expenses in the first half to simultaneously support Mech-GPT, Mech-Hand, and the iteration of existing 3D vision products — whether this sum is enough to achieve so-called "breakthrough progress" in the "embodied brain" is worth questioning.
The company's founder, Shao Tianlan, also admitted in an interview with LatePost that home scenarios and open service industries "still have no visible mature path in the short term," and that some key technologies are "not yet fully clear." It can be seen that while Mech-Mind Robot's solid foundation in industrial vision deserves recognition, mature industrial products cannot be directly equated with the still-exploratory embodied intelligence. The company derives 99% of its revenue from mature industrial 3D cameras, while embodied intelligence products such as Mech-GPT and the dexterous hand have yet to contribute substantive revenue. This means that Mech-Mind Robot's grand 30x PS narrative still lacks a corresponding revenue curve to back it up.
"Global Number One" Born in a Narrow RMB 1.8 Billion Track
For investors, the halo of "global number one" is a highly attractive label in the secondary market. Therefore, Mech-Mind Robot's "global number one" status adds another bold stroke to its valuation premium.
According to data from CIC cited in the prospectus, the AI + 3D vision-guided general intelligent robot component market in which Mech-Mind Robot operates had a total global size of only RMB 1.8 billion in 2025. Within this, Mech-Mind Robot ranked first globally with a 22.1% revenue share. However, it must be noted that Mech-Mind Robot's global number one crown sits atop an extremely small niche track — this is the background color most easily overlooked in the entire valuation logic.
Broadening the view to the more general "vision-guided robot (VGR) systems" market, the scale is much larger. According to Meticulous Research, the global VGR systems market was valued at USD 3.24 billion in 2026 and is expected to reach USD 16.92 billion by 2036, with a CAGR of 18.1%. VGR systems include 2D vision and 3D vision, covering hardware, software, and services, with application scenarios including assembly, quality inspection, and pick-and-place. The "AI + 3D vision-guided general intelligent robot components" segment where Mech-Mind Robot operates is a very small subset of the VGR market — it focuses on "3D vision guidance components" for "general intelligent robots."
Within the vision-guided robot market, Mech-Mind Robot has chosen the fastest-growing but smallest entry point. A global pie of RMB 1.8 billion means that even with a 50% share, revenue would only be RMB 900 million. Its RMB 389 million in revenue already approaches one-fifth of this niche market. Moreover, in terms of market growth rate, even though CIC's forecast is highly attractive — a CAGR of 43.2% from 2025 to 2030, with the market expanding to RMB 10.6 billion by 2030 — a horizontal comparison with the broader industry reveals that this forecast carries a notably aggressive tone. According to Interact Analysis, the global machine vision market's CAGR over the same period is only 7.2%, and China's machine vision market CAGR is about 10.1%. In other words, the forecast growth rate for Mech-Mind Robot's track is six times the global average and more than four times the domestic average.
Furthermore, this high-growth forecast is tied to a strong precondition not controlled by Mech-Mind Robot: the large-scale commercialization of general humanoid robots from 2025 to 2030, driving 3D vision component penetration from 5.1% to 10.6%. Once the pace of humanoid robot industrialization slows, the growth expectations for the entire niche market would be significantly revised downward, and the underlying logic supporting the high valuation would loosen accordingly.
A big fish in a small pond is, after all, still just a big fish in a small pond. This means that Mech-Mind Robot being a leader in a narrow track — no matter how fast it grows — cannot outpace the arrival of its ceiling. With RMB 389 million in revenue corresponding to about one-fifth of the market share, the incremental space left for the "global number one" is visibly cramped. And the high-growth narrative supporting the valuation imagination is tied to the hypothetical suspense of humanoid robot commercialization, with self-evident risks.
Conclusion
So, is Mech-Mind Robot's 30x PS overvalued? Based on the IPO price, Mech-Mind Robot's market capitalization is approximately HK$12.7 billion, corresponding to 2025 revenue of RMB 389 million, giving a price-to-sales ratio exceeding 30x. The valuation escalation process is worth noting: the company's Pre-IPO round post-investment valuation was RMB 6.367 billion, corresponding to about 16x PS on 2025 revenue; the listing market capitalization was approximately RMB 11.6 billion, with the valuation rising 82% in just two months. The prospectus did not provide sufficient explanation for the enormous valuation gap between the primary and secondary markets.
If Mech-Mind Robot's valuation level represents the premium the market pays for its global leadership, high gross margin structure, and the imaginative space of embodied intelligence, then it must be noted that Mech-Mind Robot's 30x PS is already far above its peers. In contrast, mature industrial vision companies: Cognex has a PS of about 9.5x, Opt has 9.4x, and Keyence only 6 to 7x. A simple calculation shows that Mech-Mind Robot's revenue CAGR from 2023 to 2025 was 46.6%. If the valuation reverts to the industrial vision industry norm of 9-10x PS, corresponding to 2025 revenue of RMB 389 million, the reasonable market capitalization range would be only RMB 3.5-3.9 billion — a significant premium compared to the current market value.
Of course, forward estimates cannot be directly equated with current valuations, but this comparison is sufficient to illustrate: 30x PS implies multiple optimistic assumptions — sustained high revenue growth over the long term, Mech-GPT and the dexterous hand commercializing on schedule to open new growth curves, and ultimately achieving scaled profitability. If any single assumption falls short of expectations, there is room for valuation downgrade.
The market has a simple belief that high growth rates should command high valuations. But the Mech-Mind Robot case reminds us that growth rates cannot be priced independently of track space. High-speed growth in a narrow track can easily hit the ceiling quickly. The broken IPO on the first day of listing, with the share price now at a discount of over 21%, may well be a signal of the market's rational return.