ComfyUI adds support for the MOSS-Image open-weight model
By Steven Van ·
OpenMOSS’s 5B open-weight image model renders legible text and takes structured layout prompts, and now runs in ComfyUI on consumer hardware.
ComfyUI added support for MOSS-Image on October 19, an open-weight image model from OpenMOSS. It is a 5B model that renders legible text and accepts structured layout prompts.
ComfyUI says the Comfy team worked to make the model run smoothly on any consumer hardware, while the community has already been making finetunes and LoRAs for it.
Sources: ComfyUI’s announcement and demo.