The body is commodity. The brain is everything.
From the banks of the Huangpu River, Jack Ma's Ant Group has introduced a humanoid robot into a contest that is less about metal and motors than about which nation will most deeply embed artificial intelligence into the rhythms of daily life. The R1, unveiled in Shanghai this week, reflects a broader conviction that the decisive advantage in robotics belongs not to those who build the best body, but to those who cultivate the most capable mind. In a world already reshaped by digital payments and cloud services, Ant is wagering that the next frontier of human-machine intimacy will be physical — a robot in the kitchen, the clinic, the care home.
- China's humanoid robotics race is accelerating fast enough that a fintech giant best known for processing payments felt compelled to enter the arena with a walking, cooking, medicine-sorting machine.
- The real tension is not hardware versus hardware — it is which AI 'brain' will prove intelligent enough to handle the unpredictable messiness of real homes, hospitals, and restaurants.
- Ant is threading a strategic needle: rather than manufacturing its own chassis, it is sourcing components from Chinese suppliers and backing startups, concentrating its resources on the large language model BaiLing and cloud-based reasoning systems.
- The R1 is currently confined to controlled testing in community centers and restaurants, held back by safety concerns and the absence of a commercial price — ambition outpacing readiness.
- The longer arc points toward companion and caregiver robots, machines designed to support aging populations — a market that could redefine what it means to grow old in an AI-saturated society.
Jack Ma's Ant Group unveiled a humanoid robot called the R1 at Shanghai's 2025 Inclusion Conference on the Bund, marking the fintech giant's entry into a global race already crowded with competitors like Tesla. Demonstrated by Ant's robotics unit Robbyant, the machine can guide visitors, sort medications, offer medical advice, and assist with basic cooking — a breadth of capability that signals serious intent.
What distinguishes Ant's approach is its emphasis on software over hardware. While rivals invest heavily in perfecting joints and motors, Ant is building its competitive edge around BaiLing, its own large language model, and cloud-based AI systems that allow robots to plan and execute complex, adaptive sequences of tasks. Robbyant CEO Zhu Xing envisions household humanoids not as appliances following rigid scripts, but as extensions of powerful AI — learning and deciding as they work. For a company built on digital payments, robotics is simply another channel through which to weave AI into everyday life.
China already leads the world in factory robot density, and is now pushing humanoids into less structured environments — homes, care facilities, restaurants. The R1 draws on Chinese suppliers for its physical form, with Ant-backed Galaxea AI providing the chassis and Ti5 supplying joint modules. Potential partnerships with Unitree Robotics and Orbbec are reportedly in discussion, though unconfirmed.
The robot's spatial reasoning — understanding how objects relate and how tools work in sequence — is what allows it to adapt to new recipes or unfamiliar environments, separating genuine intelligence from mere automation. Researchers caution that building a robust, scalable AI model remains the true bottleneck, even as China's scale and capital allow it to move with unusual speed.
For now, the R1 is a proof of concept, tested quietly in community centers and restaurants while Ant works through safety considerations and pricing. The longer ambition is companionship and caregiving — robots that can reshape how people age and manage health at home. Ant, long synonymous with Alipay, is making clear it intends to compete in the AI era not only through financial software, but through physical machines that carry its intelligence into the world.
Jack Ma's Ant Group stepped into the humanoid robotics arena this week, unveiling a machine called the R1 at a Shanghai conference and signaling that China's tech giants are moving decisively into a field already crowded with American competitors like Tesla. The robot, demonstrated by Ant's robotics unit Robbyant at the 2025 Inclusion Conference on the Bund, can guide visitors through spaces, sort medications in pharmacies, offer medical advice, and handle basic cooking tasks—a range of capabilities that positions it squarely in the middle of an accelerating global race.
Ant's entry matters partly because of what the company is choosing to emphasize. While other robotics makers focus on perfecting the physical machine—the joints, the motors, the body itself—Ant is betting that the real competitive advantage lies in the software brain. The company is developing its own large language model called BaiLing and using cloud-based AI systems to give its robots the ability to plan and execute complex sequences of tasks. According to Zhu Xing, the CEO of Robbyant, the vision is straightforward: humanoids in homes won't just perform chores, they'll function as extensions of powerful AI systems, making decisions and learning as they work. For Ant, a company built on digital payments and financial services, the logic is clear—robotics becomes another channel through which to embed AI into daily life.
China already leads the world in robot density on factory floors, with more machines per human worker than either the United States or Japan. The country is now preparing to move humanoids into more complex, less structured environments—homes, hospitals, restaurants, care facilities. Ant's R1 uses components from Chinese suppliers: Ti5 provides the joint modules, and Galaxea AI (which Ant backs) supplies the chassis. The company is also in talks with Unitree Robotics and Orbbec, a Shanghai-listed firm, though neither Ant nor its potential partners confirmed these discussions.
The robot's AI system allows it to understand spatial relationships—how a table relates to an appliance, how different tools function in sequence—which means it can theoretically learn new recipes and adapt to different cooking methods. This kind of reasoning is what separates a machine that follows rigid scripts from one that can handle genuine variation in the real world. Andy Mok, a researcher at the Center for China and Globalization in Beijing, noted that while manufacturing can be outsourced relatively easily, developing a robust and scalable AI model is the genuine bottleneck. China has serious competitors in this space, he cautioned, but the country also has the scale and investment capital to move fast.
Ant is not yet selling the R1 to consumers. The company is testing it in community care centers and restaurants, partly out of safety concerns, and has not announced a price. The longer-term ambition is clearer: develop robots designed to be companions and caregivers, machines that can assist with medical needs and household tasks in ways that reshape how people live as they age or face health challenges. For now, the R1 remains a proof of concept, a signal that Ant—best known for running Alipay, China's dominant digital payments platform—is serious about competing in the AI era not just through financial software but through physical machines that embody the intelligence Ant is building.
Notable Quotes
If humanoid robots are used in homes, they won't just help with everyday tasks, they'll act like super-smart brains, tapping into cloud-based AI to assist with even more things.— Zhu Xing, CEO of Robbyant
The actual physical manufacturing can probably be outsourced relatively easily, but developing a sound and scalable model is key. One of the most important prerequisites will be the AI model and not necessarily the manufacturing or engineering.— Andy Mok, Center for China and Globalization
The Hearth Conversation Another angle on the story
Why does Ant Group, a fintech company, care about building robots at all?
Because they see robots as the next interface for AI. Ant's whole business is about embedding itself into daily transactions and services. A robot in your home is just another way to deliver AI assistance—and to collect data about how people live.
But they're not building the robot hardware themselves, are they?
No. They're buying components from Chinese suppliers and focusing on the brain—the AI model that makes the robot actually useful. That's where they think the real value is. Anyone can assemble parts; not everyone can make a machine that learns and reasons.
How does this compare to what Tesla or other Western companies are doing?
Tesla is obsessed with the hardware—making the body elegant and efficient. Ant is saying: the body is commodity. Give us the software layer, and we'll make it do things competitors can't imagine yet.
Is the R1 actually ready for people to buy?
Not yet. They're testing it in restaurants and care centers. They haven't set a price. It's a demonstration of capability, not a product you can order today.
What's the real competition here—is it China versus the US, or is it just companies racing to own the robotics market?
Both. But the framing matters. China is moving faster on commercialization because it has fewer regulatory hurdles and more willingness to deploy robots in real settings. The US is still debating safety. China is learning by doing.
What happens if Ant's AI model actually works well in a robot body?
Then they've proven that you don't need to be a hardware specialist to dominate robotics. You just need better software. That changes the entire industry structure.