JD.com and Meituan have entered the same battlefield

Author | Zhou Zhiyu

Editor | Zhang Xiaoling

At the table of embodied intelligence, Meituan has always been the most aggressive player. It has long extended its reach into unmanned delivery vehicles and drones, and with the momentum of "buying half of the track," it has embraced a number of star robotics companies, including Yushu Technology.

This momentum was broken on July 21. JD.com announced its formal entry into the market with substantial investments. In one day, it announced leading investments in three top robotics companies: Qianxun Intelligent, LimX Dynamics, and Cothink Robotics.

JD.com stated to Wall Street Insights that it will announce the collaboration of its embodied intelligence brand JoyInside with many robotics companies at the WAIC to be held at the end of July.

The confrontation between JD.com and Meituan is just the tip of the iceberg. In this "promised land," regarded as the ultimate form of AI, a grander game has already begun, with tech giants like Alibaba and Tencent all in attendance, each calculating differently based on their own genes and strategies.

The influx of giants is driven by profound industrial logic: e-commerce, food delivery, and local life—these titans of the digital economy will ultimately dock at the physical ports of warehousing, logistics, and delivery. Embodied intelligence is the key to mastering this physical world and is a battleground for driving the efficiency revolution and building future core competitiveness.

As capital and strategy advance hand in hand, this technological revolution driven by top tech companies will also reshape the industrial landscape. The story of this trillion-level competition has just begun.

JD.com's "Blitzkrieg"

JD.com's moves resemble a meticulously choreographed drama. By announcing three major investments on the same day, the signal it conveys goes far beyond the funds themselves.

On July 21, JD.com's investment announcements were so dense they were almost suffocating: leading a nearly 600 million yuan Pre-A+ round financing for Qianxun Intelligent (Spirit AI); strategically leading a new round of financing for LimX Dynamics; and leading the A1 round financing for Cothink Robotics.

This was a public "muscle show," clearly intended to demonstrate its undeniable determination and strength in the embodied intelligence track.

If we break down these three companies, JD.com's strategic intent becomes clear. This is more like an "ability acquisition" targeting the core capabilities needed for future robots, rather than a simple company combination.

Qianxun Intelligent is regarded as the "academic elite" in the field. Founder Han Fengtai was a co-founder and CTO of the industrial robotics company Lashi, while co-founder Gao Yang is an assistant professor at Tsinghua University, representing a typical "scientist + industry veteran" golden combination.

Qianxun's core weapon is its self-developed VLA (Vision-Language-Action) large model Spirit v1, as well as the high-performance humanoid robot Moz1 equipped with this model. Investing in Qianxun is JD.com's bet on the most critical "brain" for its robotics legion, acquiring top-level AI capabilities that enable machines to achieve generalization and complete complex tasks.

Founded by academic heavyweight and professor at Southern University of Science and Technology Zhang Wei, LimX Dynamics is a leader in the field of motion intelligence. The company focuses on full-size general humanoid robots and bipedal robots, with its core advantage being the provision of stable, flexible, and efficient mobility capabilities for robots Its full-size humanoid robot is ready for mass production and is planned to be publicly sold in the second half of this year.

This investment is JD.com equipping the robot with reliable "legs," ensuring it can navigate stably in complex physical environments, such as JD.com's vast warehouses and changing delivery routes. The cooperation announcement from both parties also clearly states that they will "deepen collaborative efforts in retail, logistics, and services."

Industry veteran Zhao Tongyang, leading Zhongqing Robotics, holds a trump card—extreme cost control. The SE01 humanoid robot developed by them has reduced costs to within 100,000 RMB, far below similar products on the market. The key to achieving this lies in their strong self-research capabilities for core components, such as joint modules, which account for a significant portion of costs. Investing in Zhongqing is JD.com paving the "road" for the large-scale application of robots, ensuring that this cutting-edge technology is not only advanced but also economically feasible, capable of truly moving towards industrialization.

Zhongqing also stated that with the completion of two rounds of financing, they will fully accelerate their development pace in the second half of the year. In terms of products, they will vigorously promote the scale trial production and delivery process, increasing the scale and capability of the existing production and delivery team by five times; in terms of technological competitiveness, they will continue to increase investment in the research and development of embodied intelligence technology, accelerating the scenario-based implementation and commercialization of core technologies.

From these three robotics companies, it can be seen that JD.com's combination strategy is not about single-point breakthroughs but rather a full-stack layout, with ambitions clearly visible.

As JD.com repeatedly emphasized in response to Wall Street News, the investment aims to focus on supply chain scenarios and promote innovative applications.

JD.com is not just an investor but also the most eager and direct ultimate customer in this field. This dual identity of "being the largest buyer and a well-funded investor" gives JD.com significant bargaining power and strategic initiative at the negotiating table.

Building a Robot Army

If external investment is the first step in JD.com’s blueprint for building robots, then its internal organizational transformation and platform strategy are injecting soul into this "robot army," clarifying the command system, and ensuring that this grand strategy can be effectively executed.

Insiders at JD.com stated that in addition to the intensive investment in robotics, JD.com has also established relevant business departments internally, with multiple teams exploring the feasibility of implementation along different routes.

In July of this year, JD.com officially launched a platform called "JoyInside." This name cleverly pays homage to the classic "Intel Inside," and its ambition is self-evident: to become the intelligent core of the robotic era.

The core of "JoyInside" is to drive conversational AI powered by JD.com's Yanxi large model, like a pluggable "smart module," into a wide range of third-party hardware, including robots, robotic dogs, AI toys, and more.

The core task of this platform is to promote the "deep integration of robots and large models" and explore new application scenarios. It acts as a command center, efficiently coordinating and integrating JD.com's scattered resources across various business lines—technical accumulation of the Yanxi large model, scenario data from JD Logistics, and sales channels from JD Mall At the same time, JD.com has not given up on accumulating hardware capabilities. In the 2025 CVPR Robot Dual-Arm Operation Simulation Challenge, the JD Technology team defeated many well-known companies and universities to win the championship, showcasing its strong capabilities in model and data dimensions.

Under the JoyInside platform, JD.com does not need to manufacture every "body" itself but can quickly build a large hardware ecosystem by providing standardized "souls" (AI capabilities). This greatly lowers the R&D threshold for hardware manufacturers, allowing their products to become "usable" more quickly. Currently, JD.com has partnered with over ten leading robot brands, including Yuanluobo, Zhongqing, and Lingtong Robotics.

The PM01 model of Zhongqing Robotics, which received investment, has been labeled as the "JD JoyInside Robot." This model maximizes the coverage and influence of its AI technology, representing a typical platform approach.

More decisive than the platform is the organization. In mid-March this year, Liu Qiangdong appeared in Hong Kong and intensively inspected several embodied intelligence projects. According to sources from Wall Street News, Liu Qiangdong brought along two vice president-level personnel, one of whom was He Tian, the head of R&D technology at JD Logistics, and the other was Shen Hui, the head of the embodied intelligence project. Their inquiries during the inspections focused on the application of AI in real-world scenarios.

Shortly thereafter, JD Retail also established a "Smart Robotics Business" department aimed at comprehensively laying out the strategic emerging industry of smart robotics. This move is significant, marking the upgrade of the robotics business within JD.com from an exploratory project to a core strategic business unit.

At this point, JD.com's complete strategic picture in the field of embodied intelligence has become clear.

Through strategic investments, JD.com locks in cutting-edge hardware technologies and core capabilities from leading companies like Qianxun, Zhujie, and Zhongqing, ensuring it does not fall behind in hardware; centered around the "JoyInside" platform, it builds a standardized AI operating system to empower a wide range of ecological partners, achieving large-scale replication of software and intelligence; relying on the newly established smart robotics business department, it vigorously promotes the application of technology in JD.com's core businesses (logistics, retail) and pushes mature products to the market through e-commerce channels.

This approach allows JD.com to position itself at every stage of the entire value chain. It is both a technology provider and a platform operator; both an application customer and a sales channel.

This all-encompassing layout exposes JD.com's ultimate ambition: in the upcoming era of automated commerce, it not only wants to participate but also aims to quickly seize the discourse power of industry standards through open cooperation.

The "Ice" and "Fire" of the Trillion-Dollar Track

As giants wave their checkbooks and rush into the track, the entire embodied intelligence industry presents a scene of "fire" and prosperity. According to institutional estimates, this will be a market with over 300 million units by 2050, with an upstream and downstream industrial chain worth up to 1.7 trillion USD.

However, beneath this clamor, the cold reality of the difficulty of commercial landing remains a test that every entrant must face The enthusiasm in the capital market can be described as "frenzy." Yushu completed its share reform this year and officially submitted its listing guidance filing on July 18. Meanwhile, Zhiyuan Robotics plans to acquire 63.62% of the shares of Shangwei New Materials, cleverly opening up financing channels. Other embodied intelligence companies are also accelerating their financing pace, launching a sprint towards the capital market. A competition for IPO bell-ringing has already begun.

The "fire" of financing cannot melt the hard "ice" of commercialization. High costs, bottlenecks in core technology, and a lack of large-scale, verifiable commercial application scenarios are common dilemmas faced by the entire industry.

Wang Feili, an analyst in the Chinese machinery industry at UBS Securities, told Wall Street Insights that in the next three to five years, or even five to ten years, complete machine manufacturers will need to invest heavily in research and development, as well as in establishing the hardware, software, and entire ecosystem, and jointly develop new products with downstream customers. This poses relatively high financial requirements for complete machine manufacturers.

Wang Feili stated that it is still very difficult to determine who will become the final winner.

So, where is the "vitality"? The industry generally believes that before the dream of robots entering thousands of households is realized, industrial manufacturing, especially automobile manufacturing, will be the first scenario for large-scale application. This is a direction that may not be very glamorous but can generate real income and refine technology.

UBTECH's current strategic focus has significantly shifted towards the industrial sector. Its Walker S humanoid robot has entered factories of several car companies, including BYD and Nio, for practical training, performing tasks such as quality inspection, labeling, and handling.

Startups have also begun pragmatic explorations. "Everything is scene and cost-oriented," said Zhang Miao, co-founder of Lingbao CASBOT, capturing the key point. Their strategy is to "reverse engineer" the technology optimization path through real scene demands, strengthening the robot's advantageous capabilities in specific industrial scenarios while compressing costs through optimized structural design.

Looking ahead, how will the landscape of the embodied intelligence track evolve?

One possibility is a "game of giants." Giants like JD.com and Meituan, leveraging their vast proprietary application scenarios and strong capital, are building vertically integrated "walled gardens." They can internally absorb high R&D costs and prove the value of robots by improving their operational efficiency, without rushing to achieve profitability in the public market.

During this year's 618 shopping festival, JD.com's self-operated sales of intelligent robots increased threefold year-on-year, while sales of embodied intelligent robots surged 17 times. This clearly indicates that JD.com is both an application of technology and a sales channel for products, and this dual identity grants it unparalleled strategic patience and internal circulation capability.

What is the way out for independent startups? Their fate seems to have only two paths: either being acquired by a giant at the right time, becoming part of its ecological map; or finding a sufficiently vertical and deep niche market outside the core business of the giants.

All roads to Rome are filled with uncertainty. This trillion-level track competition has just begun. As the giants' robot legions have already assembled and the IPO drums of startups have begun to sound, it is still too early to conclude who can ultimately traverse the intertwined "Song of Ice and Fire" of financing frenzy and commercialization dilemmas to reach the finish line of victory The ultimate winner will inevitably be the player who can best bridge the significant gap between capital speculation and tomorrow's reality, quickly finding opportunities to create value in real scenarios