IcedCoffeeMix

https://civitai.com/models/83538

About this version

Pretty big update, but for strange reasons. I was messing around merging some of my older models into ICM 5.0 and was able to find a really good combination with ICM 5.0 and ICM 3.0, I also subsequently merged in the Add Sharpness lora I have linked and it also boosted quality by quite a bit.

A few notes:

-Faces in general seem to be improved, eyes especially.

-It's MUCH easier to generate humans without anthro features while still being ridiculously easy to prompt furries, add "human, skin" to the negative prompt and "furry, species" to the positive prompt. If you want a human just prompt without the furry tags in the positive prompt and add "furry, anthro, animal ears" to the negative.

So far it's been very versatile, it can do most species in a variety of styles. I highly recommend using it with the add detail Lora (not required, but think of it like a slider between cartoon to 2.5D).

Example prompt for the Fennec Fox in a leather jacket:

Prompt: A beautiful female fennec fox, Labyrinth, (aged brown leather jacket, canvas pants, baggy white undershirt, black collar:1.2) (detailed face, detailed eyes:1.2), stunning gaze, short wavy hair, (Stylized anime artwork, masterpiece, best quality, absurdres, P.A. Works style, solo) by sligarthetiger, by meesh, by avante92, by ashblueflames, by djpuppeh, by personalami, by kenket, by desubox, by gideon, by taran fiddler, by chunie lora:add_detail:-1

Negative prompt: blurry, out of focus, worst quality, bad-hands-5, by bad-artist-anime (bad anatomy, extra limbs, extra ears, bad hands:1.2)

Size: 800x1200

Hires upscale: 2

Steps: 35

Second pass: True

Hires Steps: 20

Sampler: DPM++ 2M

Hires Sampler: Euler a

Sampler options: rescale beta

CFG scale: 6

Clip skip: 2

CFG rescale: 0.7

Second Pass CFG scale: 6

Hires force: True

Hires upscaler: RealESRGAN 4x+ Anime6B

Face restoration: Face HiRes, Denoising strength: 0.3

HDR clamp: 0.95/2.6

Seed: 2111518298

FreeU: b1=1.2 b2=1.4 s1=0.9 s2=0.2

Hypertile UNet: 384

Please note that I am using SDNext for my images, so your results will probably not be like mine unless you are using SDNext with the exact same settings as me, I utilize FreeU and HDR Clamp and Center which both affect the final output. If you're not getting the same results as me, that is probably why.