- Deal deepens e-commerce giant’s robotics ambitions amid commercialization race
- Partnership spans co-development, scenarios and ecosystem building
Embodied AI and robotics startup MagicLab (魔法原子) has signed a comprehensive strategic partnership with e-commerce giant JD.com, aiming to jointly drive 1 billion yuan ($147 million) in sales of its full robot product lineup on the e-commerce platform over the cooperation period, Chinese media reported on June 24.
The deal comes as China’s robotics sector accelerates its shift from lab prototypes to commercial deployment.
The agreement covers joint R&D, product co-creation, scenario expansion and ecosystem development, according to the two companies.
A Dreame ecosystem player
MagicLab, founded in January 2024 in Suzhou and incubated by Dreame Technology (追觅科技), develops a full-stack robotics portfolio spanning core components, humanoid robots, quadruped robots, dexterous hands and embodied AI software platforms.
Its high-dynamic biped humanoid robot MagicBot Z1 previously made its debut exclusively on JD.com in 2025.
MagicLab said the collaboration is intended to bridge a long-standing gap between technological prototypes and large-scale commercial deployment, arguing that “purely relying on technical breakthroughs is not enough” to bring humanoid and quadruped robots into mass markets.
Speeding up commercialization
JD.com, meanwhile, has been steadily scaling its robotics strategy. In August 2025, it launched an “intelligent robotics acceleration program,” pledging more than 10 billion yuan in resources and setting a target of helping 100 robotics brands achieve over 1 billion yuan in sales within three years.
Robotics firms including Unitree (宇树科技), Agibot (智元机器人), Tiangong (天工机器人) and EngineAI (众擎机器人) have already joined JD’s ecosystem.
In March 2026, JD upgraded the program at the AWE exhibition, aiming to generate hundreds of billions of yuan in cumulative revenue for partners within the year and opening up retail, logistics, technology and healthcare scenarios for large-scale procurement and deployment.
Industry momentum is also accelerating. IDC estimates that the global intelligent robotics hardware market will approach $30 billion by 2026, with China’s embodied AI segment exceeding $11 billion.
Global humanoid robot shipments rose around 508% year-on-year in 2025, with China expected to see application scenarios expand to roughly three times current levels by 2026.
Gap between expectations and pricing
Still, the sector faces significant hurdles, including cost, reliability and real-world task performance.
Much of the industry is still caught up in a “high prototype, low mass production” phase and grappling with a gap between pricing and market expectations.
MagicLab said its products are now deployed in more than 50 countries and regions, while JD brings a global supply chain and logistics network to their collaboration.
Whether the partnership can convert technological progress into scalable commercial adoption, however, remains to be seen, analysts say.
Expanding robotics drive
For JD, however, its broader robotics drive extends beyond promotional or manufacturing partnerships.
In March 2026, embodied AI firm Spirit AI (千寻智能) signed a four-year partnership with the company, with its Moz robot deployed in JD Mall stores to perform teleoperated tasks such as coffee-making while collecting real-world interaction data.
In April 2026, JD also unveiled a full-stack embodied AI data infrastructure covering data collection, storage, labeling, training, evaluation, simulation and testing.
600,000-person data capture
The rollout came alongside its in-house wearable device JoyEgoCam. The company plans to involve up to 600,000 participants in data collection efforts.
By May, JD and the city of Suqian, the hometown of JD founder Richard Liu, had jointly launched China’s first embodied AI data collection community.
Under this initiative, residents in the northern Jiangsu city will contribute training data through everyday household activities such as cleaning and folding clothes while wearing capture devices.
