Skip to content
ComfyUI Wiki · 工作流与开源创作分享·· 28 days agoPicksAI score72

MOSS-Image's first instruction-editing LoRA released, with native support merged into ComfyUI the same day

Image editing

AI briefing

On October 22, 2026, MOSS-Image got its first instruction-editing LoRA, chinoll/Moss-Image-Edit, a rank-64 adapter that takes 1 to 4 reference images and a text instruction and covers changes to color, clothing, lighting, season, and visual style.

Why it matters

The community adds an instruction-editing LoRA for MOSS-Image; the article gives two integration paths—ComfyUI native and a node pack—along with default parameters, which can be evaluated against existing editing workflows.

Open original comfyui-wiki.com