MOSS-Image Canny Control LoRA

MOSS-Image Canny Control LoRA

Keeps the outline of a photo while the prompt changes its materials, colours and style, as a Canny edge control LoRA for MOSS-Image-SFT in ComfyUI, for people who edit and restyle images.

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 Canny edge control LoRA for the MOSS-Image-SFT image model. You give it a white-on-black edge map, and it follows the contours and framing while your prompt sets materials, colour and style. It is an edit-style LoRA, not the official MOSS colour or depth adapter. The author picked the checkpoint from training step 4500.

Who it suits: People who restyle images in ComfyUI and want the outline to stay fixed will find it fits. Designers who study layouts with a set silhouette can use it too.

Industry | Who uses it | How they use it | Example
Design | Visual designers | Recolour a composition while keeping its outline | A new colour scheme on the same layout
Product imagery | Marketers | Turn a photo's edges into another material and style | A shape rendered in a new finish
Illustration | Artists | Keep a drawn silhouette and let the prompt fill in the rest | A figure redrawn in different clothing
Research | Image-model tinkerers | Test edge control on MOSS-Image-SFT | Comparing thresholds and strengths

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
Recolour a composition | Designers | A new image with the same outline
Change material and style | Artists | The same shape in a different finish
Layout studies | Art directors | Variations where the silhouette stays put

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 Canny Edge node from comfyui_controlnet_aux
• 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-canny-controlnet

How to use it: Make a Canny map from your 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
Canny Edge node | Builds the white-on-black edge map, with low 100 and high 200 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 load a photo, build its Canny map and write a prompt for a circus ringmaster in a scarlet coat in front of a painted tent. The result keeps the outline from the photo, while the prompt supplies the clothing, colour and setting.

For newcomers and experts: A Canny map is a picture made only of edges. 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. If edges look too rigid, lower the strength toward 0.7 or raise the Canny thresholds.

You load a photo into the workflow
The Canny 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
• Edge maps with fine, tidy white lines on a black ground give the best results
• 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: 458
• Likes: 3
• License: apache-2.0
• Languages: en
• Based on: OpenMOSS-Team/MOSS-Image-SFT
• Tags: diffusers, moss-image, controlnet, lora, comfyui, canny, image-to-image

Links

• Model: https://huggingface.co/thedeoxen/MOSS-Image-canny-controlnet
• Files: https://huggingface.co/thedeoxen/MOSS-Image-canny-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-canny-controlnet

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