At the beginning of each image is of course the correct command (prompt). Especially with animals, the AI models are already very well developed. The results are almost frighteningly realistic. Already with few attributes great results can be expected. So e.g. only "Wild cat in the forest" plus an attribute like "Award Winning Photo" and already the basis for an exciting process starts.
In many cases, editing is not even necessary to create a nice light and color palette. Unfortunately, image size is the most significant problem. Properly printable, these images are only at very small sizes. For animals, however, upscaling can help, and the photos are usually detailed enough.