Quaintly Chaotic Comtesse (QCC) is my attempt at a fully open recipe merged Illustrious model. QCC is quite a dynamic checkpoint that plays well with most loras. QCC merges the features and capabilities of a collection of models that I love.

What Tools are Used to Create QCC?

The main tool used to blend the source models is Chattiori Model Merger. The final mix is then trimmed.

What Models were Merged to Create QCC?

IlluQuaint 0.8 by me
Chaos v10 by @_Kav_
Analog Madness XL5 by @CornmeisterNL
Comtesse Raffinée v1.0 by @Chumple

Merge Recipe

python merge.py WS "C:\ModelPath" "IlluQuaint_v0.8_202601040952.safetensors" "chaos_v10.safetensors" --alpha "0.5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.5, 0.5, 0.7, 0.8, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0" --output "IlluChaos_Aggressive_WS" --save_safetensors --save_half --prune
Using ComfyUI, bake in HandPatch (Unet-Only) and HD Eyes loras into checkpoint produced in Step 1.
python merge.py SPRSE "C:\ModelPath" "IlluChaos_Aggressive_WS-hands_and_eyes_patch.safetensors" "analogMadnessSDXL_xl5.safetensors" --alpha "0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.4, 0.2, 0.4, 0.6, 0.8, 1.0, 1.0, 1.0, 1.0, 1.0" --output "IlluChaos_SPRSE_Hybrid" --save_safetensors --save_half --prune
python merge.py FREQ "C:\ModelPath" "IlluChaos_SPRSE_Hybrid.safetensors" "comtesseRaffinE_v10.safetensors" --alpha "0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.4, 0.6, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8" --output "IlluChaosComtesse_FREQ_Hybrid" --save_safetensors --save_half --prune

Additional Info

CLIP skip: -2
VAE and CLIPs baked in
Practically all Schedulers and Samplers supported by Illustrious will work with QCC
CFG: 3 to 6. Steps: 20 - 40. This is just a simple guideline, please experiment by all means
Works very well with DMD2 Speed Lora (CFG: 1 - 1.5, Steps 8 - 12; again feel free to experiment)

I think that's it. What are you waiting for? Go generate some cool images with Quaintly Chaotic Comtesse and share it in the gallery!