So, the model features FlynnRider (either manly and handsome versions), Pascal the chameleon and Maximus the white horse!
The model has been trained with original images and high res images from Midjourney for realistic style. So the two main triggers for styles are (original style, 3d disney anime) and (realistic). For realistic style, it's better to use sided style lora until the v2, or other Pony checkpoint.
Some backgrounds are trained, as (the tower) or (inside the tower) and the village.