モデル/Maximilla - V1.0

Maximilla - V1.0

9/6/2025

10:30:16 PM

推奨プロンプト

score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, [][][][][] m0use_maximilla, cute, [][][][][] (solo), blue dress, (cute:1.5), upper body shot, (close up), (low angle view:1.5), indoors, happy expression, (side view:1.5), village, look at viewer, carrying potion bottles, shelves, bottles, potions

推奨ネガティブプロンプト

(giant, giantess:1.5)

score_6, score_5, score_4, (loli, cub, young, child, chibi:1.5), busty, ugly face, low res, blurry face, ugly hands, pumped body, black and white, penis, breasts, ((hair)), nipples, belly, extra ears, human

推奨パラメータ

samplers

Euler a

steps

cfg

3.5

ヒント

Fur and eye colour can be changed to adapt the LoRA for generic mouse or other rodent characters.

Model also works surprisingly well for rabbits.

With proper prompts specifying features, this LoRA can be used to generate other animals, though it may add unwanted mouse features.

Adding '(giant, giantess:1.5)' to negative prompts helps fix issues with character size relative to backgrounds.

The gallery for this LoRA is restricted to R and below to keep it wholesome.

Be aware of known issues such as incorrect rear views, bias towards happy expressions, large paws, and occasional human proportions especially in non-usual outfits.

Generates a female anthropomorphic mouse with grey fur and green eyes in a cartoon style.

Her name is Maximilla, or just Max to her friends. She is the owner and operator of the village apothecary and partner to Goldie.

Edit: I'm retroactively upgrading this to be version 1.0. Struggling to get a working LoRA for Goldie has given me some perspective on how surprisingly usable and well behaved this LoRA actually is!

Some useful notes:

-Fur and eye colour can be changed so this can work as a generic mouse LoRA.

-Seems to also works for other rodents.

-Surprisingly, it also works for rabbits.

-Other animals are sometimes possible depending on your checkpoint and the other LoRAs you are using. You might end up getting unwanted mouse features (usually ears) but if you insist on the correct features in your prompt (e.g. 'cat, cat ears, cat tail) you can usually force the LoRA to do what you want.

Some known issues:

-views from behind just look wrong 90% of the time

-strongly bias toward giving her a happy expression

-her hands/paws are often unusually large

-often makes her too large relative to background objects (adding '(giant, giantess:1.5)' to the negatives can help fix this)

-sometimes gives her (cartoon) human proportions, especially with outfits other than her usual blue dress

-Is very relucent to produce modern styles of dresses (other clothing seems to work fine). This is somewhat intentional but it's going a bit far.

Note: The gallery for the LoRA is restricted to R and below. I haven't disabled NSFW generation and I don't mind if you use the LoRA for that, I just want to keep the gallery (relatively) wholesome.