DreamToon is a simple merge (.666/.337 ratio) of the DreamBrush (https://civarchive.com/models/1537890/dream-brush) and GDPToon (https://civarchive.com/models/2191811?modelVersionId=2710136) models. It is intended to produce clean, detailed character images using simple prompts.

All showcase images use Lazy Embeddings (https://civarchive.com/models/1302719/lazy-embeddings-for-all-illustrious-noobai-pony-sdxl-models-lazypositive-lazynegative-positive-and-negative-plus-more) and basic subject tags. Most use only a character tag; some required minor descriptive prompting (i.e. Rogue). Model steers clear of inappropriate content on the whole, but occasionally gets hornier than intended (certain characters seem to bring it out-- looking at you, Velma).

Images are produced using the following params:

DPM 2M ++ Sampler

30 Steps

Resolution: 1024 x 1024

CFG 5

ADetailer for face and eyes (60% detection confidence/ 28-30% inpainting strength)

For legal purposes, this model follows and holds the same licensing agreements as its parent models.

That said, I don't believe in the licensing of AI models, or that the licensing thereof is at all morally ethical or legally binding. As such, I genuinely don't care what you use my model for-- make your money, kids. If this thing produces work people want to buy, sell it to them.