GPT Image 2.5 prompt turns cities into ramen bowls
A structured prompt for GPT Image 2.5 that builds Shanghai, New York and other cities as edible dioramas inside a ramen bowl.
Edible city prompt. More detailed than ever with GPT Image 2.5 2x2 grid, 16:9, 4 iconic cities<instructions> Input = Landscape Type (e.g., famous cities like Shanghai, Mecca, New York, Dubai ... ) Act as a Ramen Chef and Landscape Artist. Function Noodle_World ($ Landscape) Anchor: `[Bowl of Ramen] :: [Resembling $ Landscape]::5` Morphology: `Culinary diorama, a landscape of [$Landscape] constructed inside a ramen bowl, noodles forming the terrain/paths, broth forming the lakes/rivers, edible food ingredients forming skyscrapers, buildings and monuments, toppings (pork, egg, nori) forming the mountains/trees::5` Material Physics: `Oily broth sheen, steam rising, texture of cooked noodles, appetizing food colors matching the biome (e.g. red spicy broth for volcano)::3` Illumination: `Top-down food lighting, ring light reflection on broth, steam mist::2` Render Stack: `Food photography, 8k, overhead shot, Ghibli food aesthetic (but photoreal)::1` Negative: `[Real dirt, rocks, inedible, plastic bowl, dirty, cold, non-food buildings]:: -1` Output: Generate prompt for [$Landscape]. </instructions>