Value-Driven New Cycle Accelerates Xiaomi (01810) into High-End and Physical AI Dual-Wheel Drive

date
09:57 26/03/2026
avatar
GMT Eight
The Xiaomi AI strategy has entered a brand new stage of "from vision narrative to large-scale application".
Entering 2026, the global capital market's investment logic in the AI industry is undergoing a silent shift from "vision narrative" to "effectiveness verification." Compared to the abundant imagination of the grand future in 2024-2025, the current market is more focused on the clear path of technology implementation, business closure, and financial return. In this context, Xiaomi (01810) with its scarce strategic positioning advantage in the Chinese AI ecosystem - spanning technology layer (big models, chips, operating system), application layer (Miclaw/Miloco), hardware layer (phones/IoT/cars/Siasun Robot&Automation), manufacturing layer (own factories) - continues to attract the market's attention. Xiaomi's deep layout in critical segments of the industry chain has transformed it from just a consumer electronics or automotive company to a key sample in observing whether "AI deeply integrates with the real economy" can establish a viable business model. Xiaomi's performance has exceeded the scope of just financial data and has become a touchstone for evaluating the commercial viability of a full-stack AI model. On March 24th, Xiaomi released its performance announcement for 2025, achieving its strongest performance in history and providing a strong response to the market's effectiveness queries. The financial report shows that Xiaomi's total revenue reached 457.3 billion RMB in 2025, a year-on-year increase of 25.0%; adjusted net profit reached 39.2 billion RMB, a significant increase of 43.8%. Behind the record high revenue and profits is the fundamental shift in Xiaomi's growth momentum: automobiles and AI have transitioned from strategic investments to real growth engines. Crucially, entering 2026, Xiaomi's accelerated advancement in multiple fronts such as MiMo big models, Miloco smart homes, CyberOne humanoid Siasun Robot&Automation signifies that Xiaomi's AI strategy has entered a new stage of "from vision narrative to scalable application." With these developments, Xiaomi's message to the market is clear: as automobiles and AI become new growth engines, with continuous R&D investment and accelerated AI deployment in the physical world, Xiaomi, with its strategic ecosystem advantage, is poised to experience accelerated growth in intrinsic value. Based on Xiaomi's impressive financial report demonstrating a solid fundamental base and the growth potential in the new era of physical AI, several brokerage firms and investment banks have expressed optimism about Xiaomi's future development. Goldman Sachs, Citigroup, ChinaLin, among others, have given Xiaomi a "buy" rating, with target prices of 41 HKD, 42 HKD, and 44 HKD respectively. With reference to the opening price of Xiaomi's stock on March 26th, there is still over 20% upside potential in Xiaomi's stock price. Comprehensive shift in growth momentum, breakthrough in high-end strategic positions Looking at Xiaomi's 2025 financial report, the synergistic development of various business segments has collectively driven the overall performance to new highs. The IoT and consumer products, internet services segments have shown steady performance, contributing to stable cash flow and profits, enhancing the group's risk resilience. During the reporting period, the number of connected devices on the AIoT platform increased by 19.3% to reach 1.079 billion units, driving the IoT and consumer products business to achieve revenue of 123.2 billion RMB, a year-on-year increase of 18.3%. Internet services, benefiting from the overseas business performance and achieving a global monthly active user base of 754 million, recorded annual revenue of 37.4 billion RMB, a year-on-year growth of 9.7%. If the above two segments have solidified Xiaomi's steady growth foundation, then innovative businesses like smartphones, automobiles, and AI represent the "highlight" of Xiaomi's historical performance. These segments have shown significant progress in Xiaomi's high-end strategy, consolidating the brand's premium image. In 2025, China's smartphone shipments were around 285 million units, a slight decrease of 3.3% compared to the previous year. In this competitive landscape, Xiaomi's high-end smartphones have made remarkable progress. Data shows that in 2025, Xiaomi's share of smartphones priced above 3000 RMB in China reached 27.1%, a historical high; the share of smartphones priced between 4000-6000 RMB reached 17.3%, continuing to increase. This has led to Xiaomi's domestic smartphone market share hitting a record high in the high-end segment. In addition, overseas high-end sales have also reached new heights: Xiaomi's high-end smartphone sales in Western Europe grew by about 40% in 2025, with the Xiaomi T series Pro accounting for over 55% of sales; the Xiaomi 17 Ultra Leica version in Europe was priced higher than the 1T version of the iPhone 17 PM and remained popular. The recognition of high-end products in the global market has driven Xiaomi's smartphone business steadily forward in 2025, achieving annual revenue of 186.4 billion RMB, with shipments ranking among the top three globally for five consecutive years. Compared to the steady progress in the smartphone business, Xiaomi's performance in the automotive sector in 2025 can be described as "phenomenal." It has swiftly evolved from a newcomer in the industry to a dominant player, showcasing a powerful combination of "double hit models and rapid profitability," demonstrating strong product definition, ecosystem synergy, and operational efficiency, entering a new phase of "scalable growth." In 2025, Xiaomi delivered over 410,000 vehicles throughout the year, a significant increase of 200.4% year-on-year; with an average price increase to 250,000 RMB, it drove intelligent electric vehicle and innovative business revenue over the billion mark for the first time, reaching 106.1 billion RMB, a growth of 223.8%, and achieving annual operating profit. By the fourth quarter, this segment's revenue contribution had risen to 32%, approaching the smartphone business, becoming the group's strongest growth engine. As Xiaomi's vehicle deliveries expand, its high-end momentum has been fully unleashed: the new generation SU7 had over 30,000 orders within 3 days of launch and the first supercar concept car surprised the global market. Xiaomi and GT jointly released the Xiaomi Vision Gran Turismo, receiving high attention at the WMC. Xiaomi's supercar concept car is the first Chinese brand invited by the well-known simulation driving platform GT to design a vehicle, signifying Xiaomi's recognition in addition to sales volume, innovation, and brand potential, establishing its high-end presence in the international market. Firm commitment to long-term investment, accelerating AI deployment from multiple dimensions to the physical world If the comprehensive switch in growth engines is the main highlight of Xiaomi's 2025 financial report, the ongoing investment in research and development is the strategic foundation that strengthens its long-term competitive barriers and supports the company's stable fundamentals and sustained growth. In 2025, Xiaomi invested 33.1 billion RMB in research and development, close to annual profits, a growth of 37.8% year-on-year. The total number of R&D personnel reached 25,457, a 20.1% increase year-on-year. According to the latest financial disclosure, in the past five years (2021-2025), Xiaomi has cumulatively invested 105.5 billion RMB in research and development. In the future, Xiaomi will continue to implement long-termism, increasing its investment to 200 billion RMB over the next five years starting in 2026, with at least 60 billion RMB allocated to AI field in the coming three years. Continued R&D investment is propelling Xiaomi's AI achievements towards full acceleration and application, transitioning AI narratives from concepts to demonstrable and producible products, which has been prominently reflected since 2026. The standout example is Xiaomi's AI base model capabilities that have broken into the top tier globally. Recently, Xiaomi unveiled three self-developed large models aimed at the Agent era, including Xiaomi MiMo-V2-Pro. The Xiaomi MiMo-V2-Pro is Xiaomi's first trillion-parameter MoE model designed specifically for the Agent era. During the initial testing phase, the Xiaomi MiMo-V2-Pro, after being released on the global API platform OpenRouter, secured the top spot in the weekly call volume rankings in less than a week. Supported by robust AI base model capabilities, Xiaomi's "people-car-home full ecosystem" is demonstrating a unique competitive advantage, injecting stronger growth momentum into the company's full-line business. The system-level AI intelligence Miclaw embedded in mobile phones has the potential to be a key differentiating factor in Xiaomi's AI phone competition. In March of this year, Xiaomi officially launched its mobile-end Agent product Xiaomi miclaw and initiated a limited closed test. Compared to other AI execution solutions in the market, Xiaomi Miclaw's approach is fundamentally different. Whether it was the earlier exploration with the "Bean Bag AI phone" in an independent hardware form or the "Lobster" (OpenClaw) as an open-source framework, both development efforts faced challenges related to the ecosystem or security compliance. The Bean Bag AI phone faced ban issues due to direct access to system permissions and cross-app automated operations, affecting super apps' traffic and business closure; while the Lobster as an open-source tool mainly encountered fragmentation and security risks. Xiaomi's unique advantage lies in Miclaw operating as a native system app, with deep access to the phone's underlying capabilities and permissions, enabling it to explore the practical application of large models from "conversation interaction" to "system-level execution." In terms of operating paths, Xiaomi has encapsulated over 50 system-level capabilities (such as setting adjustments, permission management, cross-app data retrieval) into structured tools, dynamically called and orchestrated by the MiMo big model according to user intent. This transition from AI answering questions to completing tasks signifies a paradigm shift, establishing an essential barrier rooted in Xiaomi's full-stack capabilities of "chip-system-AI-ecosystem" that can't be replicated by any third-party application or collaborative approach. It is expected that Xiaomi Miclaw, with its unique advantage of running as a system-native application, will have a stunning impact upon public release, redefining the standards for AI smartphones and solidifying Xiaomi's leading position in the next-generation intelligent ecosystem competition. Huaxing Securities stated that the commercialized launch of miclaw on Xiaomi's flagship series 17 models would significantly enhance Xiaomi's product structure optimization and average selling price, partially offsetting storage cost pressures. They expect Xiaomi's smartphone gross margin to bottom out in the third and fourth quarters of 2026 and accelerate recovery in 2027. Considering the operational leverage effects that investors have not yet fully priced in, this would significant profit growth potential and substantial profit margin recovery. Xiaomi's exploration of AI capabilities extends beyond smartphones. In February of this year, at the MWC conference, Xiaomi showcased its future smart home exploration solution Xiaomi Miloco overseas for the first time. As the first system-level application of MiMo big models in real-life spaces, Miloco aims to extend AI from the smartphone screen to the entire living space, transitioning from "individual intelligence" to "environmental intelligence." Its implementation signifies Xiaomi's AI capabilities fully integrating into users' physical living spaces. This advancement holds strategic significance for Xiaomi's IoT business. Miloco's launch indicates that the core competitive advantage of Xiaomi's IoT business is shifting from "connecting devices at scale and cost" to "understanding scenarios and providing intelligent services." This is not just a product iteration but an evolution in business models. When AI is successfully integrated into living spaces through Miloco, Xiaomi's IoT business moves away from the linear growth logic of hardware sales towards enhancing single-user ecosystem value, expanding software services, strengthening full-stack technological barriers, and achieving exponential value reappraisal. The intelligent evolution of the IoT business is not isolated but relies on Xiaomi's vast user ecosystem and interaction gateways. With over 160 million monthly active users, Xiao Ai, a critical strategic asset for Xiaomi in the Agent era, has stored data, scenarios, user habits, and commercial capabilities, a core resource encompassing the four elements. As industry competition shifts from a contest on model parameters to practical application deployment and ecosystem construction, leveraging this asset, Xiaomi has taken the lead in converting technology into scalable commercial value. This not only proves Xiaomi's success in the past AIoT strategy but also becomes the solid cornerstone for its entry into the new era of the "ecological intelligence entity." Simultaneously, Xiaomi's AI capability layout has begun to systematically accelerate its extension into the mobile space. During Xiaomi's spring new product launch on March 19th, Xiaomi's end-to-end assisted driving system for cars witnessed a groundbreaking architecture-level innovation powered by large modelsXiaomi XLA cognition big model was officially released. With core features of "more modes, higher efficiency, and greater control," it propels intelligent driving from traditional "data-driven" to a "cognition-driven" paradigm. This transformation will fundamentally reshape the relationship between humans and cars, transitioning vehicles from complex tools that users need to learn and adapt to into "intelligent companions" that can proactively understand the environment, anticipate needs, and seamlessly integrate into users' daily lives. The technology operates in the background, while the experience moves to the front, truly achieving a shift from "people adapting to cars" to "cars adapting to people," enhancing the overall user experience. This is not just a technological iteration but a critical move for Xiaomi in building long-term differentiation barriers in the intelligent automobile field. Moreover, Xiaomi has achieved substantial breakthroughs in its humanoid Siasun Robot & Automation business. In March of this year, Xiaomi's humanoid Siasun Robot & Automation officially entered Xiaomi's car factory and began a "work experience" on a real production line - autonomously running for 3 hours at the self-tapping screw mounting station. It achieved a dual-side mounting success rate of 90.2%, fully matching the 76-second production line rhythm requirements. This milestone signals a crucial step towards the operational application of Xiaomi's scaled industrial automation. This marks a qualitative shift from "technology showcase" to "productivity verification" and embodies Xiaomi's "AI + hardware + manufacturing" full-stack capabilities. Real-world production line data proves that the humanoid Siasun Robot & Automation has preliminary industrial practical value. With the ongoing deployment towards scalability, Xiaomi is set to be the first to successfully run humanoid Siasun Robot & Automation in an industrial setting, building a complete business loop and accumulating invaluable practical experience and data barriers for further entry into complex environments such as households, services, and beyond. With the impending outbreak of physical AI, Xiaomi is set to become an essential core target in the new era that cannot be bypassed The competitive landscape of AI applications is fundamentally shifting. While most existing applications remain in the realm of processing text, images, and code, the true industrial revolution and value explosion will come from the new era of AI perceiving, understanding, and directly manipulating the physical world - covering fields from Siasun Robot & Automation, autonomous driving to intelligent environments with autonomous execution capabilities. In this transition from the digital to the physical realm, mere superiority in technical parameters is not the decisive factor in consumer markets. The true moat lies in the scale of ecosystem entry and the depth of accumulated data. Only platforms that enable high-frequency, high-quality, and trustworthy utilization of AI capabilities by a vast number of users in diverse scenarios can transform technology into sustainable commercial value. From this perspective, Xiaomi has displayed a unique strategic advantage. Its positioning has transcended singular hardware offerings, encompassing a complete "perception-decision-execution" cycle, spanning technology layer (big models, chips, operating system), application layer (Miclaw/Miloco), hardware layer (phones/IoT/cars/Siasun Robot&Automation), and manufacturing layer (own factories). This interconnected "people-car-home full ecosystem" has planted Xiaomi's AI in the real physical world and users' lives from the outset of R&D, creating a robust strategic competition barrier. Currently, a consensus has begun to emerge in the market, indicating that leaders in the AI era may not necessarily be newcomers starting from scratch, but established "giants" that have already established high-frequency, deep, and trustworthy relationships with their ecosystems and users. Because the future AI competition is not just a battle of algorithms but also a comprehensive competition of ecosystem scale, scenario data, and commercialization paths. With Xiaomi having the deepest ecosystem entry, it has the potential to nurture the most robust application-based AI in the future, positioning itself as an essential core target in the new era of physical AI.