For best outputs:

Sampler: DPM++ 3M SDE Karras

Steps: 40, cfg 7, CLIP skip 2.

VAE: sdxl_vae.safetensors.

Pos prompt: the model listens incredibly well, so keep them short and don't overuse weights.

Neg prompt: no need or what you dont want to see.

It has a diverse range of photographic styles — nature, portrait, candid, astro — and camera angles/focal lengths — telephoto, ultrawide, close, wide, macro, etc.

Because it's intended to produce images with a photographic style, prompts for things that are typically 2D (such as drawings, anime, etc.) tend to get a bit of "pop."

It's capable of making less than perfect looking people. Hands are pretty good. And it can actually crank out some usable text, which is a fun bonus.