Based on the interim results for the first half of 2026 released by GEEKPLUS-W (02590), new order intake reached RMB 2.385 billion, a year-on-year increase of 35.5%. Revenue climbed 25.3% to RMB 1.284 billion, while gross margin improved further to 35.8%, and the adjusted net loss narrowed by 32.1%. Excluding investments in embodied AI research and development, the core business is notably close to breaking even. Interpreting these figures solely as evidence of an AMR leader's "growth" may understate the significance of this interim report.

Geekplus is displaying a new growth structure: the warehouse AMR foundation continues to expand, manufacturing scenarios are breaking through rapidly, subscription-based services are gaining scale, and embodied AI has secured collaborations with Fortune Global 500 companies. In other words, the company is transitioning toward a multi-engine business system driven by products, scenarios, services, AI, and embodied intelligence working in tandem.

Orders up 35.5%, subscription services surge, and the "post-robot business" accelerates realization

When assessing a robotics company's operational health, order flow often provides a more forward-looking signal than current revenue. In H1 2026, Geekplus secured new orders worth RMB 2.385 billion, up 35.5% year-on-year. This indicates that despite an already substantial revenue base, the company continues to demonstrate strong new project acquisition capabilities, while building a solid order backlog for future revenue growth. More important than the overall order expansion is the continuous optimization of its order structure.

If hardware sales determine a robotics company's scale, then software and operational services are likely to define its future profitability ceiling. In H1 2026, new orders from subscription-based services reached RMB 156 million, growing more than 75% year-on-year, with the Americas region seeing a 455% surge. Traditional robotics projects are inherently project-based: clients invest, companies deliver, revenue is recognized, and then operations move into a maintenance phase. Subscription services, however, shift the relationship from "one-off project delivery" to "ongoing operations." As AI technology accelerates its integration into warehouse operations, robotics providers are no longer just offering software to coordinate robot movement; they are increasingly taking on inventory optimization, task allocation, operational analytics, exception management, and even overall warehouse efficiency enhancement.

This fundamentally alters the value proposition of robotics companies. In the past, clients purchased robot equipment; in the future, they may be purchasing the capability for continuously improving logistics efficiency. Geekplus's AI-enabled smart operations subscription service builds a layer of software and service capabilities that generate ongoing value on top of already deployed robots and systems. Leveraging a global network of over 1,000 clients and 81,000 deployed robots, this business can scale rapidly, offering high renewal rates, strong stickiness, and predictable revenue—a key pathway for robotics firms to enhance long-term revenue and profitability.

Tote-to-person up over 200%, manufacturing orders surge over 600%, with warehouse and manufacturing scenarios expanding in tandem

The core warehouse business is also deepening. Order growth for Geekplus's tote-to-person product line exceeded 200% year-on-year. While AMRs have traditionally been most mature in bin and shelf picking applications, rising client automation levels are driving robotic systems into full-pallet handling, high-density storage, and upstream-downstream warehouse connections, further solidifying the company's leadership in pallet warehouse automation. Meanwhile, orders from production and manufacturing scenarios surged more than 600% year-on-year. Globally, the most mature AMR demand has long been concentrated in e-commerce, retail, and warehousing fulfillment. Manufacturing production logistics, characterized by complex processes, diverse workflows, and high system integration requirements, represents a critical frontier for robotics companies seeking to broaden market boundaries. Geekplus is forging a distinctive path: entering a client's warehouse ecosystem first, then naturally extending along the client's supply chain and production processes into manufacturing logistics.

This "warehouse-to-factory" expansion model offers clear advantages over relying solely on new client acquisition. Existing system data and project delivery experience lower the decision-making cost for clients making additional purchases, while enabling product synergies and cross-selling opportunities across different scenarios within the same client. From this perspective, the over-600% growth in manufacturing orders signifies not just a new business segment, but a transition of the company's robotics capabilities from "warehouse automation" toward "intra-enterprise logistics automation."

Profitability continues to improve as scale effects begin to materialize

Rapid growth and profitability are often the two hardest challenges for robotics companies to solve simultaneously. In H1 2026, Geekplus reported an adjusted net loss of RMB 60.6 million, narrowing by 32.1% year-on-year. Notably, the company invested RMB 44.5 million in embodied AI R&D during the period; excluding this investment, the adjusted net loss was just RMB 16.1 million, a narrowing of 81.9%, bringing the core business close to breakeven. The signal behind these numbers is clear: the global scale effects built up over years of investment are progressively showing up on the income statement. For robotics companies, once a certain scale is reached, R&D platforms, core components, supply chains, software systems, and global delivery infrastructure can be shared across multiple product lines.

Geekplus's current product strategy reinforces this. Different robot products share a common technology foundation and core components, reducing new product R&D costs while enabling supply chain scale benefits and cross-selling opportunities. Its business model is gradually evolving from a "high-growth robotics company" toward a phase where "scale growth and profitability improvement proceed in parallel."

New RoboShuttle Hyper product signals further execution of the "full-scenario strategy"

Another notable shift is Geekplus's more proactive entry into high-growth niche segments. In August 2026, the company launched the RoboShuttle Hyper climbing goods-to-person solution, with a system throughput capacity of up to 6,000 bins per hour per 1,000 square meters, targeting high-volume, high-throughput applications such as MFC micro-fulfillment centers. From a product logic standpoint, this is not simply about adding another robot type. The global intelligent logistics market is entering an increasingly segmented phase. Previously, companies might rely on a relatively standard set of robots to address a wide range of problems, but as automation matures, clients are demanding more refined solutions tailored to throughput, storage density, flexibility, and payback periods. Consequently, the competitive battleground for leading robotics firms is likely shifting from who owns a "star robot" to who commands a comprehensive product matrix covering more scenarios.

Geekplus's existing product portfolio—including goods-to-person, pallet-to-person, goods-to-person, RAPS robotic arm picking stations, and the Gino1 wheeled humanoid robot—is built around this logic. Richer product offerings amplify the scale benefits of shared technology platforms, and broader scenario coverage expands cross-selling opportunities within the same client base. This explains why "full-scenario" has become a key keyword in Geekplus's next-stage strategy.

Embodied AI products secure Fortune Global 500 partnerships, targeting cumulative shipments exceeding 10,000 units in three years

If warehouse AMRs define Geekplus's current foundation, embodied AI determines the ceiling of its growth potential over the coming years. In H1 2026, the company's embodied AI business secured collaborations with multiple Forbes Fortune Global 500 companies. Geekplus plans to leverage its data flywheel and GINO ECO ecosystem strategy to drive humanoid robot-AMR collaboration in warehouse scenarios, building a robot-powered productivity system covering storage, picking, handling, and patrolling. The company also intends to extend applications into manufacturing, retail, and other scenarios over the next three years, with cumulative embodied AI product shipments targeting over 10,000 units.

Compared to the crowded field of competition centered on "humanoid robot hardware," Geekplus's approach stands out distinctly: instead of hunting for application scenarios from the lab, it derives embodied AI requirements from existing, real-world commercial use cases. This is a crucial distinction. The biggest challenge in embodied AI has never been merely completing a demo; it is whether systems can operate reliably in complex real-world environments, generate ROI that clients are willing to pay for, and scale repeatably. Warehousing and logistics happen to be among the most suitable scenarios for embodied AI's initial commercialization: tasks are relatively well-defined, environments are comparatively structured, and there is substantial demand for flexible, manual-like operations that traditional AMRs cannot fully cover. Importantly, Geekplus does not require humanoid robots to autonomously handle entire warehouse automation. In its vision, humanoid robots and AMRs will work in synergy: AMRs handle mature mobility, handling, and transport tasks, while humanoid robots take on dexterous manipulation work, all coordinated through a unified software and AI system. This implies that the future competitive unit may no longer be "a single robot," but rather an intelligent productivity system composed of a collective of robots. If this pathway proves viable, Geekplus's accumulated AMR technology, global client base, warehouse data, and software systems could all become foundational infrastructure for embodied AI commercialization—another reason its embodied AI strategy warrants long-term industry attention.

From "AMR champion" to "embodied AI platform company," the commercial boundary continues to widen

Reviewing Geekplus's H1 2026 performance, several clear trends emerge. First, the core AMR business continues to demonstrate robust growth, providing a stable foundation. Second, subscription services are converting a vast installed base into a steady stream of service revenue, translating into high-repeat, high-margin, sustainable income. Third, the global client network and full-scenario strategy are accelerating scenario expansion and cross-selling capabilities. Fourth, scale effects are consistently materializing, bringing the core business close to breakeven. Fifth, embodied AI is transitioning from a phase of technology investment into real-world commercialization. Together, these five threads point to a single transformation: Geekplus is no longer a traditional "AMR supplier," but a "smart robotics platform company" that enters with innovative robotic solutions, maintains long-term engagement through AI software and intelligent operations, and is powered by general-purpose embodied intelligence as its long-term technology engine.

This trajectory carries representative significance for the broader Chinese robotics industry. In the first phase, Chinese robotics companies proved they could build globally competitive products. In the second phase, they needed to demonstrate entry into global mainstream client supply chains. The third phase truly tests global client operations, ongoing service capabilities, AI prowess, and platform-level capabilities that transcend individual robot form factors. Geekplus is entering this third phase. The RMB 2.385 billion in new orders, overseas revenue share exceeding 75%, manufacturing order growth surpassing 600%, subscription services growth above 75%, and the advancement of embodied AI commercialization each stem from different business dimensions but together form a common signal: this Chinese robotics company, which has consistently ranked among the global AMR leaders for years, is unlocking a growth boundary far larger than AMR alone. For Geekplus, H1 2026 is less a "performance growth milestone" than a starting point for structural changes in its business model. The next phase worth watching is whether, as the AMR, AI smart operations, and embodied AI business curves begin to compound, the company can convert its global scale advantages into sustained growth and ultimately establish an intelligent robotics ecosystem built for the global market.