Jianzhi Robotics Secures Hundreds of Millions in Funding, Led by Ant Group and Didi

On June 1, Jianzhi Robotics, a provider of embodied intelligence solutions, announced the completion of multiple consecutive rounds of financing, raising several hundred million RMB in total. This round was jointly led by Ant Group, DiDi, and Delian Capital, with participation from existing investors including Shunwei Capital, Baidu Ventures (BV), and Jiushi Intelligence. This marks the largest single investment to date in the “body-agnostic data” segment of embodied intelligence, making Jianzhi Robotics the best-funded company in this field. Jianzhi Robotics stated that the proceeds will be primarily allocated to three key areas: (1) expanding multimodal human behavioral data and continuously upgrading its data collection product portfolio; (2) building a foundational large-model technology stack for data to enhance precision, consistency, and completeness; and (3) establishing an end-to-end closed-loop training infrastructure for embodied models by integrating the full pipeline—from data production and model training adaptation to performance evaluation. The company also plans to accelerate its global expansion and deepen ecosystem partnerships across industries. Focusing on high-fidelity human behavioral data captured in real-world scenarios, Jianzhi Robotics has pioneered a “model-defined data standard” approach and built a multimodal data acquisition system encompassing vision, touch, audio, and more. Its proprietary visual modules and wireless communication technologies enable multi-device synchronization with latency under 1 millisecond. The company also introduced the industry’s first “multi-camera perception matrix,” composed of six 200-megapixel RGB cameras, enabling full-field, high-precision recording of environmental and behavioral information. Additionally, Jianzhi Robotics has developed an integrated data capture architecture combining Ego (first-person perspective), Fingers (hand manipulation), and high-precision tactile sensing. Specifically, Gen DAS Ego captures first-person visual input and head pose trajectories; Gen DAS Fingers replicates fine-grained human hand movements through a biomimetic structure; and the accompanying tactile module simultaneously records implicit data such as contact force and dynamic feedback. Together, these three components enable comprehensive recording of human-robot interaction processes in real-world settings. To date, Jianzhi Robotics has established a specialized embodied intelligence data production line, engaging over 3,000 data contributors and accumulating more than one million hours of data assets across over 10,000 real-world environments—including homes, factories, commercial spaces, logistics centers, laboratories, and medical facilities—covering more than 2,000 everyday human operational skills. Its Ego+Fingers head-hand coordinated high-precision data collection capacity now exceeds 100,000 hours per month, and the company has forged deep collaborations with more than 30 leading AI enterprises worldwide.