36Kr reports on developments involving AgiBot

36Kr reported on developments involving AgiBot on October 29, 2026.

In evaluating robots, correlations in simulation, rankings and demonstrations alone cannot confirm operation in real-world settings. Reports involving AgiBot should also be read separately from materials that show specific operating conditions and track records.

The first distinction to make when reading this report is between technical evaluations of robots, simulation performance, demonstrations and operational track records in real-world settings. These are not equivalent measures, and results in one area cannot be used to infer achievements at another stage. The subject indicated by the headline should also be checked separately from materials that substantiate specific technologies or operations. See the original sources for specific claims and supporting evidence. Related coverage by this newspaper reported that AgiBot's simulation system supports simulation of deformable objects and showed score correlations with real robots across four types of robots and 22 tasks AgiBot updates Genie Sim, reports simulation of deformable objects and correlation with real robots. This report concerns the relationship between simulation and real robots; it is not evidence of AgiBot deployments or operations. Another article covering an embodied intelligence evaluation ranking led by the University of Hong Kong treated rankings, data and collective movement as distinct evaluation criteria University of Hong Kong-led embodied intelligence evaluation ranking reported by leaderobot.com. It is important not to interpret evaluation results or the existence of datasets as evidence of sustained operations in the field. The same applies to trade shows. ROBOTECH 2026 opens in Shenzhen, with exhibits ranging from robots and components to secondary development reported on exhibits including robots, components and secondary development, but the presence of exhibits alone cannot confirm results in real-world settings. The focus of future verification will therefore be whether public information specifies the robots and tasks involved, operating environments, continuity of operations and methods for measuring results. The information provided here does not confirm these specific conditions, so no judgment can be made about the scale or results of AgiBot's actual deployments.