MOSS-Image Scribble Control LoRA

MOSS-Image Scribble Control LoRA

Turns loose white strokes on black into a finished image while the prompt sets texture and style, as a scribble control LoRA for MOSS-Image-SFT in ComfyUI, for people who start from rough drawings.

Made by thedeoxen on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: This is a scribble control LoRA for the MOSS-Image-SFT image model. You give it loose white strokes on a black ground, and it keeps to your rough drawing while the prompt supplies the look. It is an edit-style LoRA, not the official MOSS colour or depth adapter. The author picked the checkpoint from training step 5000.

Who it suits: People who sketch a composition by hand and want a finished image from it will find it fits. Designers who plan a scene with only broad shapes can use it too.

Industry | Who uses it | How they use it | Example
Illustration | Artists | Draw a few strokes and let the prompt fill in the rest | A quick figure sketch turned into a styled portrait
Film and storyboards | Storyboard artists | Turn rough panels into finished frames | A three-stroke street scene rendered as a still
Design | Art directors | Test layouts before committing to detail | Several looks tried on one composition
Research | Image-model tinkerers | Test scribble control on MOSS-Image-SFT | Comparing strengths and stroke styles

If you do not use ComfyUI or MOSS-Image-SFT, you can skip it, since it is trained for that base model.

Where to use it:

• ComfyUI, with the workflow file included in the repository
• ComfyUI's stock node for encoding MOSS edit prompts
• Windows, macOS or Linux with Python, to download the files

Use case | Who | What they get back
Draw a composition | Artists | A finished image that keeps your layout
Storyboard to frame | Storyboard artists | A styled still made from rough strokes
Quick layout test | Art directors | Several looks on the same shapes

How to set it up: You need ComfyUI, the MOSS-Image-SFT files and this LoRA. The author names three base files and a workflow.

• ComfyUI with the soft-edge node that the included workflow uses
• MOSS-Image-SFT: the UNet moss_image_sft_int8_convrot.safetensors
• A Qwen3.5 text encoder file for MOSS-Image, with the moss_image type chosen when you load it
• The VAE flux2-vae_small.safetensors

1. Install the Hugging Face hub package.
2. Download this repository's files, which hold the LoRA and the workflow.
3. Load the MOSS-Image-SFT UNet, text encoder and VAE.
4. Apply this LoRA to the UNet with a strength of 1.0.

pip install -U huggingface_hub
hf download thedeoxen/MOSS-Image-scribble-controlnet

How to use it: Draw white strokes on black, or build a map from a photo, then hand it to the Text Encode MOSS-Image Edit node as the only reference image. On the positive and the negative encode alike, switch the edit format to ai-toolkit. Sample with 30 steps, CFG 4, and euler with simple. There is no trigger word.

Tool | What it does
Soft-edge node | Builds the white-stroke map from a photo, with safe mode on in the workflow
Text Encode MOSS-Image Edit | Takes the control image and your prompt
LoRA loader | Applies this LoRA at strength 1.0
Included workflow | Loads a photo, builds the map, encodes, and samples

Example: You draw a few loose white strokes for a standing figure on a black ground. You write a prompt for a circus ringmaster in a scarlet coat in front of a painted tent. The result keeps the pose and layout, while the prompt supplies the clothing, colour and setting.

For newcomers and experts: A scribble map is a handful of loose white lines on a black picture. A good first step is to run the included workflow on one photo with the default settings, then change only the prompt.

Its adapter weights use rank 64 for the linear layers and rank 16 for the convolutions. It learned as an edit adapter. The control picture serves as the starting reference, and the text describes the image you want out. It was trained with the kv cache on, which is why the edit format is ai-toolkit. The training strokes were thicker and softer than Canny or lineart lines, so soft-edge maps match them better than a thin sketch map. If the control feels too rigid, lower the strength toward 0.7.

You draw strokes or load a photo into the workflow
The soft-edge node builds a white-on-black map
The map and your prompt go to the encode node
MOSS-Image-SFT with the LoRA samples the new image

Good to know:

• License: apache-2.0, as listed on Hugging Face
• Cost to run: none named
• Data it reaches: the images and prompts you load into ComfyUI, and the source names no web or account access
• Sparse, bold strokes on a black ground suit it better than a busy map
• Not for: other base models, since it was trained on and is recommended with MOSS-Image-SFT
• It is not the official MOSS colour or depth adapter

On Hugging Face, 13 November 2026

• Task: image to image
• Library: diffusers
• Downloads in the last 30 days: 377
• Likes: 3
• License: apache-2.0
• Languages: en
• Based on: OpenMOSS-Team/MOSS-Image-SFT
• Tags: diffusers, moss-image, controlnet, lora, comfyui, scribble, image-to-image

Links

• Model: https://huggingface.co/thedeoxen/MOSS-Image-scribble-controlnet
• Files: https://huggingface.co/thedeoxen/MOSS-Image-scribble-controlnet/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download thedeoxen/MOSS-Image-scribble-controlnet

Then follow the usage in its model card. Nothing in this kit runs by itself.