Chinese robotics company X Square Robot has completed four consecutive financing rounds, including a Series C investment, valuing the business at more than $2.8 billion, or 20 billion yuan. The Shenzhen-based company said the funding will support research into embodied artificial intelligence foundation models, commercial deployments and robotics infrastructure. It did not disclose the amount raised in the individual rounds or the total capital secured.
IDG participated in the Series C round, while venture capital firm HSG and electronics manufacturer Xiaomi invested in multiple earlier rounds. X Square Robot has also received lead investments at various stages from Meituan, Alibaba and ByteDance.
The company said it is the only embodied AI business in China to have secured lead-round backing from Meituan, Alibaba, ByteDance and Xiaomi. That claim could not be independently verified from the information provided.
Founded in 2023, X Square Robot develops systems that combine AI foundation models, robotics hardware and data-processing infrastructure. Its technology is intended to help robots perceive their surroundings, make decisions and perform tasks in changing physical environments without relying solely on predefined rules.
Founder and Chief Executive Wang Qian said the company had focused on developing its foundation models internally, alongside systems for collecting and processing training data and testing robots in operational settings.
X Square Robot introduced WALL-B, an embodied AI foundation model based on its World Unified Model architecture, in April 2026. According to the company, the model trains perception, language, action and physical prediction capabilities within a single network rather than connecting separate vision, language and action components.
The company has also made two models, WALL-OSS-0.5 and WALL-WM, available as open-source software. It said WALL-OSS-0.5 completed more than 80% of four selected tasks from a set of 17 tests involving physical robots without additional post-training. The company did not provide comparative results or details about the testing conditions.
WALL-WM is designed to predict events by aligning language, visual and action data. X Square Robot said this approach is intended to improve cross-modal learning and predictions about physical interactions. The company has developed a data pipeline covering automated collection, cleaning, annotation, quality control and augmentation. Data generated through commercial deployments is used to refine its models and create datasets for less common or more complex situations.
X Square Robot is testing its technology in household, industrial and logistics settings. It has partnered with online marketplace operator 58.com on a robotic cleaning service in Shenzhen and Beijing, where robots work alongside human cleaners in residential properties.
The company also began a programme in May under which robots are placed in participating households for periods of up to one month. The programme is designed to test how the machines respond to a range of domestic tasks and interactions. X Square Robot said information gathered from these deployments will be used to improve model performance and support the development of robots capable of operating across multiple environments and tasks. “As AI moves beyond digital experiences into the physical world, progress will depend on close integration between models, data and robotics,” Wang said.
