Model Information
.jpeg&w=1920&q=75)
.jpeg&w=1920&q=75)
.jpeg&w=1920&q=75)
.jpeg&w=1920&q=75)
.jpeg&w=1920&q=75)
Description
Important:
You have to use epsilon sampling in your model queue or your images will be colorful noise.
Just put a ComfyUI native ModelSamplingDiscrete node in front of your KSampler and set it to eps.
Ok, so why this?
Inspired by their fuse work on CatWai, this is a merge between IllMoods v1.5, CyberRealistic Pony Catalyst v5, and NoobAI. Aiming for the best of all worlds, it is looking and responding great, although slightly differently which I like about it. It plays well with LoRAs too.
Use your fav SDXL resolutions, but I feel it does good things at 1024x1536, euler normal. Give it a try.
positive:
masterpiece, best quality, absurdresnegative:
lowres, worst quality, low quality, bad anatomy, bad hands, bad feet, 4koma, comic, greyscale, censored, jpeg artifactsIllMoods-Noobz (Illu vs NoobAI)
166downloads
Download ModelModel Details
- Type
- AI Model
- Task
- text-to-image
- Subtype
- Safetensors / Checkpoint AI Model
- Created
- Updated
- June 6, 2026