I've seen several video tutorials and it appears like your AI is able to maintain the original product or reference image, but I've also seen a few people in their recent reviews that have raised concern about the generated image 'altering' or 'changing' the original product.

This seems to be a common problem with AI generated images, the way it would change text to something random and gibberish or alter the design/pattern of products. However, I know some of the latest models have been getting a whole lot better with consistency between images, especially with the recent Flux Kontext model.

Does your AI now use Flux Kontext to maximize product consistency across image generations?

This is very important as maintaining the original product/design is the main use-case of your tool, so this needs to be perfected.