On March 8th, Xiaohongshu's Super Intelligence team introduced the latest iteration of its image editing model, FireRed-Image-Edit 1.1. This release comes just under a month after the launch of version 1.0. The team highlights that this new model boasts significantly enhanced capabilities in several key areas, including ID-consistent editing (ensuring the subject's identity remains unchanged during edits), multi-element fusion (combining multiple visual elements seamlessly), portrait makeup (applying cosmetic enhancements to portraits), and font style referencing (mimicking specific font styles). Additionally, the model supports a wide range of training and deployment optimizations, achieving remarkable efficiency with only 4.5 seconds of processing time and a mere 30GB of VRAM usage. To foster transparency and collaboration within the tech community, the project code, technical report, model parameters, and a comprehensive training-distillation-inference framework have all been made openly available.