There are simpler ways of doing it. Vore games like "from where you live" use sprites against a consistent background. The sprites are reusable and lead to a faster production process. My visual novel does not use sprites and I generate and edit each image used, with some images only having small edits (facial expressions) or flip back and forth during conversations. There are currently 588 unique images in the visual novel. At 50 prompts per image - roughly 29,400 prompts over 6 months (only including images actually used.)