For best outputs and how to use:

Sampler: Restart

Steps: 30, cfg 7, CLIP skip 2

VAE: sdxl_vae.safetensors

This model has a trigger, added usually at the beginning of the pos: "score_9, score_8_up, score_7_up, best quality, masterpiece, 4k, uncensored, prefect lighting, rating_explicit, very aesthetic, anime BREAK". Explained by the creator:

"This model recommends using "score_9, score_8_up, score_7_up, best quality, masterpiece, 4k, uncensored, prefect lighting, rating_explicit, very aesthetic, anime BREAK" as a starting.

It is because ponyDiffusionV6XL did not release the artist list, so We don’t know the supported artist styles"

Also, this model requires of this LoRa, "neg4all_bdsqlsz_xl_V7" at a weight of 0.9 (included in the galleries)

For the neg, nothing or the one included in my galleries. This embedding, "negativeXLD", also included in the galleries, is mandatory.

The model recognizes most of the common poses, and with a skilled prompt, can show what u want. Feel free to use my galleries as example of how to tailor the pos prompt.

This model is able to reach certain level of semi-realism, or also to slightly understand other cartoon styles but anime, like pixar, 3d or disney. To add realism, just ask it, like "realistic, hyperrealistic, real life, cinematic" ect.