Version 1.0 is an experimental mix of Bismuth and Wicked illustrious models embelished with some LORAs. It produces highly detailed semirealistic outputs, but tends to add too much. I recommend using simple prompts with it with low CFG (gives best results in my opinion).

Version 2.0 is a complete rework using ther models as the base with a shift in direction closer to cartoon/anime style.

Version 3.0: back to 1.0 formula with improvements in stability.

Version 4.0: updated mix with the latest versions of Bismuth and Wicked illustrious mixes. (some interesting results on low cfg or steps)

Models used in the mix (for v2):

• https://civitai.com/models/1231009/be-mix-illustrious

• https://civitai.com/models/1412760?modelVersionId=1684082

• https://civitai.com/models/1162518?modelVersionId=1714314

• https://civitai.com/models/1074049?modelVersionId=1736745

LORAs used in the mix:

• https://civitai.com/models/389869?modelVersionId=1169143

• https://civitai.com/models/113055?modelVersionId=1115701

• https://civitai.com/models/1018079?modelVersionId=1141520

Recommended setitings:

Sampling method: Euler a, DPM++ 2M SDE, DPM++ 2M, DPM++ 2S a, DPM2 a, DEIS

Schedule type: Karras, SGM Uniform

Sampling steps: 20-50

CFG Scale: 3-7

Highres.fix + Adetailer for face are recommended