LoRA in the style of sambaneko. It's a silhouette vector art style.

Aside from the trigger sambanekostyle, try using simple background to start, as high as 2.0. Adjust up or down depending the on desired outcome. This will tame the background effects to a more subtle level. Adding silhouette can help to make the character blend more into the background, while omitting it can make the character more 'on top' of the background.

I made this AnimagineV31 version for people who prefer Animagine over Pony, and to work with Animagine's ecosystem of models and LoRA. Compared to the PonyXL version, while much worse at ponies, I feel in this version the background effects are more controlled, somewhat better defined, and centered around the character better.

The training data is entirely ponies (though it generalized to anime very well). An attempt was made to keep all of the pony data and tags, but if you want show accurate pony images in sambaneko's style, please check out the PonyXL Version instead as it's FAR better at creating ponies.

There's a lot of creative possibility here, so just try raising, lowering, and adding tags to see what happens. It has often reacted in ways that surprised me.

Trigger:

sambanekostyle

Recommended prompt examples:

Positive prompt:

sambanekostyle, masterpiece, best quality, very aesthetic, absurdres

solo, simple background, centered composition, {{tags}}

Negative prompt:

low quality, worst quality, normal quality