Tech Startup Eyes Cheaper Robots, New Applications
In Shanghai, a tech startup is making significant strides in the mass production of humanoid robots, with the goal of making them accessible to a broader market. AgiBot, an artificial intelligence and robotics company, is at the forefront of this effort, aiming to drastically reduce costs while expanding the potential applications of these advanced machines.
At a high-tech factory in Shanghai’s Lin-gang Special Area, engineers are assembling humanoid robots. This factory, operational for less than a year, is the first in Shanghai to mass-produce these robots, having already manufactured over 1,500 units, according to AgiBot. The company plans to open a second factory in Shanghai this year and hopes to reach a production capacity of 10,000 units annually.
“At scale, our robots will cost under 200,000 yuan ($27,500) — less expensive than a family car,” stated Peng Zhihui, co-founder of AgiBot. Production has already surged from an initial goal of 100 units per month to more than 15 units per day, driven by strong market demand. With anticipation building for 2025 as a pivotal year for humanoid robot mass production Chinese companies like AgiBot are keen to demonstrate their technological and manufacturing capabilities.
A recent list of 100 global humanoid robotics-related companies, released by Morgan Stanley, highlights China’s growing presence, with 37 companies listed, including UBTech and BYD. The China Electronics Society forecasts that China’s humanoid robot market will reach 870 billion yuan by 2030.
Humanoid robotics embodies the concept of embodied AI, a term emphasized in China’s 2025 Government Work Report, which refers to AI integrated in physical hardware. Qiao Hong, an academician at the Chinese Academy of Sciences, commented, “Humanoid robots are a particularly promising area. China has established a core technological foundation for the large-scale and cost-effective production of high-performance robotic systems.”
AgiBot attributes its rapid expansion to Shanghai’s advanced manufacturing infrastructure and the Yangtze River Delta’s robust supply chain. Zhang Shaozheng, general manager of manufacturing at AgiBot, said, “From an industrial chain perspective, we’ve leveraged synergies with the new energy sector, particularly in components like electric motors and gearboxes. It is precisely these mature supply chains that enabled us to produce humanoid robots at scale in such a short timeframe.”
Beyond the physical construction of the robots, developing their “brain”—the AI—is crucial. To accelerate robot training, AgiBot recently launched GO-1, an AI large language model. The model enables robots to quickly learn new tasks with minimal training, reducing technical barriers for embodied AI applications, AgiBot explained. The GO-1 model uses real-world human demonstrations and online video data to improve its understanding of human actions.
At AgiBot’s robot training center, 100 robots are being trained to simulate everyday tasks, such as pouring tea, folding laundry, and setting tables in environments modeled after cafes and living rooms. These robots collect high-quality action datasets by mimicking human behavior which are vital for training the GO-1 model.
Yao Maoqing, a partner of AgiBot and president of its embodied intelligence business unit, explained: “Even a simple water-pouring action requires nearly 100 high-quality data samples for training. These datasets, collected from the 100 robots, are uploaded to the cloud computing platform for model iteration, eventually enabling universal task deployment.”
The initial focus for these humanoid robots is on repetitive industrial tasks, with logistics being a prominent example. Yao highlights how these robots can adapt to challenges in packaging variable-sized items, a task that requires real-time adjustment—a task well-suited for humanoid robots. “AI control allows them to rethink strategies upon failure, much like humans,” he noted.
As the technology matures, AgiBot anticipates many broader applications. “Humanoid robots’ flexibility fills the gap between rigid automation and human labor,” Yao said.
