Has anyone been testing Grok Imagine 1.5 recently?

I've been using Grok Imagine 1.5 for marketing visuals and concept images over the past few weeks.

One thing I've noticed is that it performs much better when prompts focus on lighting, camera angle, and composition rather than just style keywords. The outputs feel more consistent, and I spend less time regenerating images.

I'm still comparing it with a few other image models, but it's becoming a regular part of my workflow. Curious to hear what prompting techniques are working well for everyone else.