Skip to content
comfyorg · YouTube 创作视频· ComfyUI·· 2026-10-20PicksAI score72

ComfyUI supports the MOSS-Image image model

MOSS-Image is now supported in ComfyUI

AI briefing

ComfyUI announces support for OpenMOSS's open-source image model MOSS-Image, a 5B-scale model that can render clear, legible text and accepts structural layout prompts. The official workflow has been released on Hugging Face, aimed at image generation scenarios requiring layout control such as posters, advertising, and localized creative work.

Why it matters

ComfyUI integrates the open-source 5B image model MOSS-Image, whose text rendering and structural layout control offer direct reference value for poster and advertising image generation.

Open original youtube.com