TMTPOST -- Alibaba Cloud is making its first move into the burgeoning embodied intelligence sector, leading a near-1 billion yuan ($140 million) funding round for Chinese robotics startup X Square Robot.

The investment, announced on Monday, marks a milestone in China’s push to develop AI-powered physical robotics and signals Alibaba Cloud’s commitment to supporting next-generation intelligent systems.

(Photo taken at the 2025 World Robot Conference)

The Series A+ round was co-led by Alibaba Cloud and CASI, with participation from China Development Financial, Sequoia China, and INCE Capital. Existing shareholder Meituan Strategic Investment made an oversubscribed follow-on investment, while Lenovo Star and Legend Capital also continued to back the company. The fresh capital will be deployed to further develop X Square Robot’s self-built general embodied intelligence models and advance its hardware product line.

What sets this deal apart is Alibaba Cloud’s debut as an investor in an embodied intelligence company. Sources say the cloud unit will provide X Square Robot with “multidimensional support” beyond financing, potentially including cloud infrastructure, AI compute capabilities, and integration with Alibaba’s broader digital ecosystem. Previously, Alibaba Group has invested in other robotics-related ventures such as Unitree Robotics, Zhujili Power, and Lingxin Qiaoshou, but this represents its first direct entry into a company focused on general-purpose embodied AI.

Founded in December 2023, X Square Robot has quickly emerged as a leader in China’s next-generation robotics landscape. Prior to this round, the company completed seven financing rounds, raising over 1 billion yuan cumulatively. Its founding team combines elite academic credentials with industry experience: CEO Wang Qian holds a PhD from the University of Southern California and both bachelor’s and master’s degrees from Tsinghua University, while CTO Wang Hao has a PhD in computational physics from Peking University and previously led large-model algorithm development at the Greater Bay Area Institute of Digital Economy.

X Square Robot’s technical approach focuses on achieving general embodied intelligence through end-to-end unified large models. The company recently unveiled its Quanta X2, a wheeled dual-arm humanoid robot capable of multimodal large-model control. The robot boasts 62 degrees of freedom, tactile sensing, and operates within a 0–2 meter workspace. On the software side, X Square Robot has developed the WALL-A series of VLA (Vision-Language-Action) operational models, integrating perception, reasoning, and action. The models have demonstrated zero-shot generalization on novel tasks, a key step toward adaptable, general-purpose robotics.

In late August, the company open-sourced its foundational model, Wall-OsS, along with training code for developers, aiming to accelerate adoption and practical applications on diverse hardware platforms.

Wang Qian, Founder and CEO of X Square Robot

Despite rapid progress, X Square Robot is prioritizing wheeled robots over bipedal designs, focusing on indoor and semi-controlled environments such as hotels and nursing homes. “Our short-term goal is to tackle commercial service and public service scenarios,” Wang said during the 2025 World Robot Conference. “Successfully completing complex tasks closer to the consumer end will highlight the reasoning advantages of our WALL-A model.”

Wang emphasized that the company’s approach integrates hardware and software, delivering complete solutions to end users. He added that general embodied large models are foundational models for the physical world, distinct from large language models, but operating in parallel with them. He projects that successful adoption in a flagship scenario could open a market worth hundreds of billions, if not trillions, of yuan.

The timing of X Square Robot’s funding coincides with a surge in China’s embodied intelligence sector. Other startups—including Unitree, Galbot, Spirit AI, Robotera, LimX Dynamics, and TARS—have each secured single rounds exceeding 500 million yuan.

Major Chinese tech firms such as JD.com, Meituan, and Ant Group are also actively investing in the sector, with JD.com backing Spirit AI, EngineAI, and LimX Dynamics, and Meituan backing Unitree, Galbot, and multiple other robotics ventures.

Globally, venture capital interest in humanoid and embodied robotics remains robust. Pitchbook data shows that more than $5 billion has been invested in humanoid robot startups since early 2024, highlighting the sector’s rapid maturation and strategic importance.

Despite the excitement, Wang cautions that widespread commercial adoption is still several years away. Unlike autonomous driving, he said, embodied intelligence will not see a “sudden leap” to maturity. He expects the industry to reach a “ChatGPT moment” in robotics within three to five years, contingent on scaling laws, sufficient data, and ongoing improvements in model architecture and training.

“Ensuring that scaling laws continue to apply is key—we have already confirmed that. Combined with larger models, more data, and ongoing training improvements, we see a clear path ahead,” Wang said.