Live Updates > Live Update Details
2026-04-09 14:33:02
[JD.com Open Sources JoyAI-Image-Edit, a Multimodal Foundation Model: Pioneering the Integration of Spatial Intelligence into a Unified Framework to Empower Embodied Intelligence] (1) JD.com Research Institute has officially open-sourced the multimodal foundation model JoyAI-Image-Edit, which supports text-to-image processing, image understanding, and instruction-guided image editing. (2) It is the industry's first open-source model to deeply integrate spatial intelligence into a unified multimodal framework. (3) In the field of embodied intelligence, this model can start from limited real images and synthesize a large number of spatially consistent multi-view training samples, providing underlying capability support for vision-language-action systems and world models.
