For SEER TECH (06106), which only debuted on the Hong Kong Stock Exchange on June 24, the significance of its first interim results goes beyond delivering a set of high-growth figures. It marks the company's transition from the technology and business model validation phase into a new stage where scale effects are gradually being unleashed. According to information obtained by Zhitong Finance APP, for the first half of 2026, SEER TECH achieved revenue of RMB 264 million, a year-on-year increase of 67.5%. Gross profit reached RMB 123 million, up 71.7% year-on-year, with the overall gross margin rising from 45.4% in the same period last year to 46.6%. The loss attributable to equity shareholders narrowed by 25.1% year-on-year to RMB 37.89 million, while the adjusted net loss narrowed by 50.1% year-on-year to RMB 10.97 million.
However, compared to the single metric of revenue growth, the comprehensive resonance of operational indicators deserves more attention. In the first half of 2026, shipments of the "Robot Brain" surpassed 8,000 units, a year-on-year increase of over 80%. New orders secured exceeded RMB 467 million, growing over 60% year-on-year. The number of new customers added surpassed 400, an increase of approximately 65% year-on-year, and the existing customer repurchase rate exceeded 60%. The mutual corroboration of these five indicators—revenue, shipments, orders, customers, and repurchase rate—suggests that SEER TECH's growth is not driven by a single large contract or periodic demand, but is built upon the combined forces of customer expansion, repeat purchases from existing clients, and industry penetration.
The Enhancement of "Brain" Capabilities is the Underlying Logic for Overall Growth
Looking at the financial report's product classification, in the first half of 2026, SEER TECH generated revenue of RMB 185 million from intelligent robots, a 70% year-on-year increase, accounting for approximately 70% of total revenue. Revenue from the "Robot Brain" reached RMB 59.02 million, up 56.1% year-on-year, representing about 22% of the total. But these two revenue streams are not isolated businesses. At its core, SEER TECH is an embodied intelligence platform company centered on the "Robot Brain," with underlying capabilities leaning towards software and algorithms.
The Robot Brain from SEER TECH is not just a single control hardware or an independent model capability. It is a tripartite architecture integrating the "Model + Cerebellum + Nerves." The model layer is responsible for environmental understanding, task reasoning, and decision planning, serving as the core of the robot's cognitive abilities. The cerebellum layer handles motion control and execution scheduling, translating decisions into precise and stable robot actions. The nerve layer connects multimodal perception with the entire machine hardware, enabling real-time collection of sensor data and efficient command dispatch. These three components operate in a closed loop, forming a complete intelligent chain from perception to action for the robot. Intelligent robots are the crucial vehicle for the brain's capabilities to enter the physical world and complete scenario-based delivery. While its revenue growth superficially appears as expanded whole-machine sales, the core driving force is the continuous enhancement of the "brain" in areas like perception, decision planning, motion control, cluster scheduling, cross-configuration adaptation, and complex task execution. As the brain becomes smarter and more stable, robots can progress from "being operational" to "accomplishing tasks," enabling large-scale deployment across more customer and industry scenarios. Therefore, the 70% year-on-year growth in intelligent robot revenue not only indicates higher market acceptance for its whole-machine products but also reflects that the company's underlying algorithms, models, and software platforms are accelerating their transformation into deliverable and replicable commercial outcomes.
The commercial value of the company's "brain" capability manifests in two ways. One is delivering the "brain" directly to customers as a standalone product (integrating the model, cerebellum, and nerves into a single piece of hardware), corresponding to the RMB 59.02 million in revenue for the reporting period. The other involves embedding the "brain" into whole machines like humanoid robots, embodied forklifts, composite robots, and delivery robots, generating intelligent robot revenue through a software-hardware integrated form. Consequently, the 22% "brain" revenue share in the financial statements does not fully reflect the "brain's" actual contribution to the company's overall revenue and product competitiveness. The company disclosed in its announcement that the "Robot Brain" has a gross margin of approximately 80%. This high added value stems not only from the controller hardware but is built upon a unified control architecture, algorithms, software interfaces, and development tools. This same underlying capability can adapt to different robot forms, component combinations, and application scenarios, helping customers lower the barriers to robot R&D, adaptation, and delivery. The growth in intelligent robot sales further expands the "brain's" installation base and application scope, allowing algorithms, toolchains, and platform capabilities to be reused across more products and customers. This growth logic, where enhanced brain capabilities boost product competitiveness and expanded whole-machine deployment broadens platform reuse, is precisely what distinguishes SEER TECH from traditional robot hardware companies.
In the first half of 2026, the company's gross profit increased by RMB 51.41 million year-on-year, already exceeding the combined year-on-year increase of approximately RMB 46.05 million in selling, administrative, and R&D expenses. As the "brain" capability continues to translate into whole-machine sales, customer coverage, and scenario replication, the company's business model, characterized by software-centric foundations and high platform reuse rates, is beginning to break through revenue growth and reflect in operational efficiency.
Overseas Revenue, Orders, and Customers Accelerate Simultaneously
If the domestic business validates SEER TECH's commercialization capabilities, the overseas business is beginning to validate the replicability of its platform. In the first half of 2026, the company's overseas revenue reached RMB 65.81 million, a year-on-year surge of 197.5%, with its revenue share rising from 14.0% to 24.9%. New overseas orders exceeded RMB 124 million, growing more than 550% year-on-year. The gross margin for the overseas business remained at a high level of approximately 67%. As of the end of June, the company's customers spanned 43 countries and regions, with the number of new overseas customers added in the first half of 2026 increasing by over 150% year-on-year. The synchronized growth in overseas revenue, orders, and customers indicates that the company's internationalization is shifting from the channel development phase to the phase of actual revenue contribution. More importantly, the overseas business not only brings revenue increment but may also improve the company's profit structure. The overseas gross margin is significantly higher than the group's overall level. As the local sales, service, and channel systems gradually improve, the overseas business is expected to become a dual driver for both revenue growth and profitability enhancement. It is important to note that revenue recognition from overseas new orders will be reflected in the coming reporting periods, subject to project progress, delivery, and acceptance.
"Platform Company Base + General Brain Option"
Another important characteristic of SEER TECH is its independence from reliance on a single robot form. According to data from CIC cited in the company's announcement, SEER TECH ranked first globally in intelligent robot brain sales for three consecutive years from 2023 to 2025, holding an approximate 25% global market share in 2025. By sales volume, the company's global ranking in the industrial intelligent robot market rose from third in 2024 to second in 2025. In a robotics industry report released on August 6, 2026, SPDB International pointed out that industrial robots have entered a phase of large-scale maturity, with growth primarily driven by domestic substitution, manufacturing automation, and exports. Humanoid robots, on the other hand, are still in the early stages of mass production, with technology roadmaps and competitive landscapes yet to be finalized. In the short term, these two are not substitutes; rather, industrial robots undertake standardized, high-precision tasks, while humanoid robots supplement flexible and unstructured environments. A CMB International industry report from August 14 also suggests that while foundational motion control for humanoid robots is progressing rapidly, the "brain's" generalization capability and real-world data remain the primary bottlenecks for large-scale commercialization. The report, citing IDC data, noted that in 2025, scenarios like commercial shows, scientific research, and data collection accounted for 78.4% of humanoid robot shipments, while truly economical industrial applications are still being validated. This implies that SEER TECH's existing business can provide a stable base of revenue and scenarios. A further breakthrough in its brain capabilities could empower humanoid robots, robot dogs, embodied forklifts, and AI delivery robots, creating new growth avenues. Compared to betting on a single hardware manufacturer or a specific technological route, a cross-configuration platform can share in the developmental dividends of multiple robot forms while diversifying risks while the landscape of hardware manufacturers remains undecided. CMB International also used Tesla's Optimus as an example for estimation, suggesting that with large-scale cost reduction of mechanical components, the overall hardware cost of a whole machine could potentially decline by nearly 70% in the medium to long term, while the relative cost share of domain controllers and chips might actually increase. This projection is an institutional scenario assumption, not industry or company guidance, but it reveals a trend: the more standardized and cheaper the robot hardware becomes, the higher the relative value of control, models, and software within the industry chain is likely to rise.
Tens of Thousands of Real-Machine Deployments Lay the Foundation for a Data and Model Closed Loop
The long-term competitiveness of an embodied intelligence enterprise lies not merely in the scale of its data reserves, but more critically in whether the data originates from real production tasks, whether it can form a continuous feedback mechanism, and whether it can effectively drive a positive iteration cycle between model capabilities and product experience. As of the end of June 2026, robots equipped with SEER TECH's "Robot Brain" have achieved cumulative deployments on a scale of tens of thousands of units, landing in over 2,500 customers across more than 20 sub-industries, including 3C electronics, automotive manufacturing, new energy, semiconductors, biopharmaceuticals, and retail. This has built a diverse and solid real-world scenario foundation for the data closed loop. Based on this large-scale scenario deployment, the company has established a full-chain embodied intelligence infrastructure (Infra), covering the entire process from data access, cleaning and desensitization, data annotation, model training, simulation evaluation, model release, and edge-side data feedback. Currently, the company's self-developed end-to-end (E2E) and Vision-Language-Action (VLA) models have been applied to the X1 PRO wheeled humanoid robot, embodied intelligent forklifts, and the P300 and P1500 AI delivery robots, achieving batch deployment among global leading enterprises in the semiconductor, automotive, and 3C industries. This system forms a sustainable positive growth flywheel: robots running in real-world scenarios generate frontline operational data, which feeds back to drive continuous model iteration. As model capabilities improve, subsequent scenario adaptation and project deployment costs are further reduced, leading to more product shipments and data accumulation, creating a self-sustaining cycle of "Deployment – Data – Model – Growth." According to a research memo from Tianfeng Electric New Energy, the company currently operates approximately 50,000 robots externally, with a data feedback ratio of about 20%. It has accumulated 500,000 hours of cross-configuration real-machine operation data, characterized by authenticity, scenario diversity, standard consistency, and sustainable iteration. Notably, the company's vertical VLA model has been successfully deployed at Nvidia's Houston factory. Based on current shipment growth rates, Tianfeng Securities projects that the company's real-machine operation data will surpass one million hours by 2027. Tianfeng Securities points out that while the general-purpose robot brain is not yet fully mature, industrial scenarios are highly structured with fewer extreme cases. Training VLA models with real-machine data is sufficient to meet deployment needs. SEER TECH's controller, as a highly standardized product, can rapidly scale in industrial scenarios while simultaneously collecting real-machine data to feed model training, forming a solid data flywheel. This foundation can then be used to train foundational large models, ultimately progressing towards the long-term goal of a general-purpose robot brain.
Re-evaluating SEER TECH: Significant Upside Potential for the Stock Price
The listing has significantly bolstered the company's financial safety cushion and strategic investment capacity. As of the end of June 2026, the company held approximately RMB 1.012 billion in cash and cash equivalents, with bank borrowings of around RMB 92.01 million, resulting in a net cash position of approximately RMB 920 million. The net proceeds of approximately HK$995 million from the global offering remained unutilized as of that date. This ample capital reserve provides solid support for the company's continued investment in areas such as embodied intelligence infrastructure, world models, computing resources, new product R&D, and global service network construction. Looking ahead, SEER TECH will continue to refine its embodied intelligence infrastructure, integrating data feedback, model training, simulation evaluation, and model infusion. It will also advance R&D in spatial intelligence, world models, and WAM to enhance robot perception, prediction, planning, and cross-scenario generalization capabilities. Building on this, the company will accelerate the productization of technology, iterating on multi-form humanoid robots and robot dogs, and launching a new generation of "Robot Brain" controllers. Through a unified platform, it aims to lower the barriers to development and application, opening up more industrial and commercial scenarios. From an investment perspective, positioning in SEER TECH is not about betting on the success or failure of a specific humanoid robot model. Instead, it's a wager on the industry's long-term, fundamental demand for a unified "brain," a general-purpose control platform, and real-world scenario data closed loops as robot forms continue to diversify. This is the core value differentiation of a platform company compared to a single hardware manufacturer.
At the valuation level, data from SPDB International as of August 4 indicates that the average forecast Price-to-Sales (PS) ratio for 2026 among industrial, mobile, service, and humanoid robot whole-machine companies is approximately 7.9 times. In contrast, core component and upstream supplier companies have an average forecast PS ratio of about 14.3 times, showing that upstream platform companies command a significant valuation premium. Calculations by several institutions also confirm the current valuation's attractiveness. One institution's calculation shows that the company's current market capitalization corresponds to only about half of its base business valuation (RMB 10.5-14 billion), translating to a 2026 PS of less than 10 times (compared to Cognex enjoying a 10 times PS even during its single-digit growth phase). Compared to valuations of RMB 20 billion and above for similar targets in the domestic primary market, and the US$10-30 billion valuation levels of overseas benchmark companies, the company's embodied intelligence "brain" option value appears nearly free. Tianfeng Securities also views the company as a scarce asset in the domestic robot brain track, projecting revenues of approximately RMB 720 million and RMB 1.12 billion for 2026 and 2027, respectively, corresponding to a PS of only 9.7 and 6.3 times. The firm believes the company's business model is more similar to Momenta's, both being "asset-light + software development" models. Considering the higher ceiling of the embodied intelligence track and the company's potential for over 50% compound growth in the future, Tianfeng Securities assigns a 15 times PS valuation, implying a 138% upside for the stock price, and maintains a "Recommend" rating. Overall, revenue, orders, customer base, deployment scale, and global layout are merely the starting point for SEER TECH's growth. Looking ahead, whether these existing advantages can be continuously translated into higher growth quality, stronger cash generation capabilities, and more solid platform barriers will be the core determinants of its long-term shareholder value and valuation center.
Comments