For the best outputs:

Sampler: DPM++ 2M SDE Karras

Steps: 20, cfg 7, CLIP skip 2.

VAE: sdxl_vae.safetensors

This model is anime focused and anime oriented, although it understands of art styles. If you name, for example, a Monet, it will try to create a mix of both.

This anime model can do from anime to something close to realism. If you want pure anime, simple put "Illustration, anime" in the pos. If want to reach semi-realism, add keywords like "Cinematic Photo", "Ultra realistic" or "RAW photograph", adding some of weight to it even. In fact, the model listens very well to all this kind of words, so if you plan to create pure anime, avoid any keyword relate to quality, volumetry, ect. Simple depict what you want to see.

In the neg: convenient to add the word "desaturated".

The model is obedient, it listens very well. If it pops twins, simply add "solo" in the pos, for example.

It's an SDXL model, keep prompts short and precise.