Booster Robotics Raises Tens of Millions of RMB in Pre-A Round Led by Source Code Capital
The proceeds from this round will go toward small-batch production and market expansion, talent recruitment, and ongoing R&D iterations.


36Kr has learned that humanoid robotics company Booster Robotics recently closed a Pre-A round of tens of millions of RMB, led by Source Code Capital with participation from Waterwood Capital and Yingang Capital. The funding will go toward small-batch production and market expansion, talent recruitment, and continued R&D iteration.
Founded in 2023, Booster Robotics builds humanoid robots and developer tools for the global developer community, aiming to drive a productivity revolution through open collaboration.
In a recent product demo, Booster Robotics showed its BR002 humanoid robot rising from a prone position by folding its legs backward — the same standing technique demonstrated by Boston Dynamics' electric Atlas.
The team was able to quickly replicate Boston Dynamics' movements because it had committed early to a human-leg configuration with a distinctive hip joint layout capable of 360° rotation — converging on the same design philosophy independently. Compared to quadruped, bird-leg, or simplified-leg designs, the human-leg configuration poses far greater challenges for motion control algorithms but offers substantially higher ceiling for locomotive capability. Booster Robotics' robot can perform both lateral and longitudinal splits, with joints that rotate a full 360°.
Hao Cheng, Booster Robotics' founder, said the company is initially focused on building the robot's cerebellum (motion control) and hardware body. The immediate priority is achieving mass production and establishing a viable business model before investing heavily in compute units and robot operating systems.
"What we're good at now is the body and cerebellum, but I believe our real moat won't be in the brain, cerebellum, or body — it will be in the operating system and compute unit. Future robots must have their own compute unit, their own SoC chip. The other piece is the [operating] system. Robots will need corresponding systems and developer tools so that other developers can build on top of them."
Gong Li, investment VP at Source Code Capital, said: "Among humanoid robotics entrants, Booster Robotics stands out as one of the rare teams with full-stack capabilities spanning hardware, algorithms, and commercialization, with more than a decade of industry experience. We believe that under Hao Cheng's leadership, Booster Robotics will continue to capture the commanding heights of humanoid technology and be among the first to achieve mass production and delivery."
Humanoid robots can be divided into the brain, the cerebellum (which governs motion control), and the hardware body. Booster Robotics is entering through the cerebellum and hardware body. Motion control for humanoids carries extremely high barriers: solving it involves substantial physics and mathematics, the algorithms are complex, proprietary configurations are scarce, and forward development is difficult. Software engineering complexity is also high, with tough decoupling challenges and numerous hurdles in layered software packaging.
Founder Hao Cheng holds bachelor's and master's degrees from Tsinghua University and previously served as VP of Product at ByteDance's Lark. Mingguo Zhao, a researcher at Tsinghua University's Department of Automation and director of the Brain-Inspired Robotics Center at the Robot Control Laboratory and Unmanned Systems Center, serves as the company's chief scientist.
Booster Robotics has assembled a rare team combining expertise in robotics, software, and commercialization, and has been product-oriented from day one.
The company plans to officially launch its humanoid robot product with an integrated motion control algorithm development platform in the second half of this year, alongside small-batch production and sales.
Previously, Booster Robotics closed an angel round of tens of millions of RMB in October 2023.
Source: 36Kr | Author: Wang Qin | Editor: Qiu Xiaofen


