Workflows / MOSS-Image
Free workflow · new
MOSS-Image
The open model that topped Civitai this month. Apache 2.0.
MOSS-Image (OpenMOSS): 5B parameters, a Qwen3.5-9B text encoder, text that reads right and 2K output, under a licence with no strings. ComfyUI's official recipe (25 steps, CFG 2.5, skip-layer guidance), plus optional 8-step DMD2 turbo and character-LoRA slots. Realism finetunes available as a pack.
Drag the .json onto ComfyUI. Runs best on the Gmanski RunPod template.
You get
- Gmanski MOSS workflow
- Turbo and character-LoRA slots (bypassed)
You need
- ComfyUI 0.42+
- 16 GB VRAM (int8)
Model packs · Text to image · Characters
mossmoss-turboOn the template: gmanski-packs install moss moss-turbo
Pairs well with.
MiniMax H3newCharacter Video Swap — one photo, any performance
A guide video gives the moves. Your photo gives the person.
MiniMax H3newH3 Image Edit — up to 8 references
Edit a picture with other pictures, in 8 steps.
Krea 2newBulk Character Swap — Dynamic LoRA
A folder of photos in. Your character in every one — and the right LoRAs picked per picture.