Milos (Crave)

Well, main thing I guess is, I follow this structure for my prompts: Subject + Action + Environment + Style + Detail + Quality. With those things in mind, you pretty much just fill in what you're wanting. I typically use keywords/tags (example: 1boy, blonde hair, blonde undercut, etc) but descriptive sentences work too. Depending on what app you use for images, results may vary. I use PixAI, which deals with a lot of different models and LoRAs, these can heavily influence the image you'll generate. I'll attach a brief example of a prompt of mine, along with the image it produced (｡•́ωก̀｡)

ngl, had to look that up. But I see it now (｡•́ωก̀｡)