Zidong Taichu Open-Sources ZDTaichu5.0-9B: A General-Purpose Multimodal Large Model Leading in Spatial Embodied Capabilities Within 10B Parameters

Zidong Taichu officially open-sources ZDTaichu5.0-9B. The model's size is suited to deployment needs, and among general-purpose multimodal large models with under 10B parameters, it stands out in spatial understanding and embodied capabilities, while maintaining a high level of general multimodal ability and fully covering the entire chain of capabilities including perception, cognition, decision-making, and action. More notably, this open-source release not only opens the model weights but also discloses a detailed data production plan, and has already been validated in real physical scenarios such as scientific research automation and intelligent manufacturing. This open-source release means that domestic multimodal large models are moving from "understanding the digital world" to "understanding and acting in the physical world."