This JSON packs a five-image series into one prompt: pizza with a long mozzarella stretch, syrup-soaked baklava, a molten lava cake, a sticky cinnamon roll and a creamy cheesecake, each torn apart by two hands with its own lighting note. When run as a single image, the model rendered only the first entry, a pizza cheese pull, and it is a strong one: blistered crust, basil leaves, steam and crumbs mid-air, with long glossy strands of cheese between slice and pie. To get the whole set, run each 'prompt' value separately with the same ratio and render mode. It's ideal for bakery, pizzeria or delivery-app social feeds that need a consistent action series.
