Image ChatGPT Image 2.5
Glossy miniature diorama of Paris with tiny chibi figures and 3D title blocks
A soft-studio-lit, tilt-shift cityscape of Paris as a layered diorama, complete with iconic landmarks, native signage, and playful chibi characters.
Prompt
2x2 grid, 16:9, do this for 4 famous cities: class CityPostcardDiorama:
def __init__(self, PLACE):
s = infer(PLACE) # title, native, icons, layers, signage, chibi, palette
self.s = s
def render(self):
return glossy_miniature(
title=blocks_3d(self.s.title) + native(self.s.native), # P₁
tableau={ cat: icon_of(PLACE, cat) for cat in ICON_CATEGORIES }, # P₂
planes=assign_layers(self.s.layers), # P₃
life=chibi(self.s.chibi), signage=vertical_native(self.s.signage), # P₄
palette=self.s.palette, light=SOFT_STUDIO, tilt_shift=True ) # P₅
def verify(self, img):
assert title_3d_dominant(img) and native_present(img), "P₁: flat/no native"
assert covered(img.tableau, ICON_CATEGORIES) and \
all(of_place(i, PLACE) for i in img.icons), "P₂: sparse/generic"
assert all(non_empty(p) for p in planes(img)), "P₃: empty plane/flat"
assert chibi_present(img) and script_of(img.signage, PLACE), "P₄: no life/lorem"
assert glossy_toy(img) and saturated(img), "P₅"
CityPostcardDiorama(PLACE).render()
Published: October 3, 2026 by Gadgetify