Truth to tell, I prefer the comic illustration style that I can get out of Z Image with just prompting (no LoRA) but this is a distinctive style I developed for SDXL and Flux and it seems appropriate to keep it alive for my new favorite model.
The model brings strong outlines and bold coloring, with a somewhat haphazard bit of cross-hatching inkwork and a tendency to make everything a flat plane if possible.
Generally, I refer to 'illustration of ...' at the start of my prompt, and Z Image stays fairly true to the style.