GPT Image Prompt
How to useCopy this prompt and paste it into ChatGPT, Gemini, Nano Banana, or your favorite AI image tool. Adjust details if needed.
Create a polished square 1:1 surreal miniature diorama artwork with cinematic magical realism.
Scene: An enormous weathered open book lies diagonally across a warm beige studio floor. The lower halves of its pages dissolve into a shallow translucent turquoise pond that floods across the paper, while the upper page areas remain dry and densely covered with non-legible antique print texture. Sand, crumbled page fibers, torn parchment edges, faint algae stains, and small wet fragments gather around the book, making the boundary between literature and landscape feel physically believable.
Main subject: One original, non-identifiable child around 7–9 years old, with feminine presentation, light neutral-toned skin, a slender child build, and chin-length tousled golden-blonde hair that partly hides the face from the high angle. She is barefoot and wears a modest long-sleeved, knee-length deep navy-blue dress with realistic fabric weight, seams, folds, and a softly flared skirt. Place her ankle-deep on the flooded right page, torso angled toward the right edge, head bowed toward the water, one arm naturally extending a long-handled round fishing net beside a small red metal bucket. Keep hands, feet, limbs, and child proportions coherent; do not reproduce any identifiable likeness.
Action and objects: Show exactly three bright orange-and-ivory koi fish swimming through the foreground and middle of the page-pond, with one subtle darker fish silhouette beneath the surface. Add one small red bucket with a curved wire handle at the far-right edge. Create gentle ripples around the child’s feet, the net, and the fish.
Composition and camera: Use a high oblique three-quarter overhead viewpoint with a 50mm macro-diorama look. Let the open book occupy roughly 80% of the frame, running from the lower-left foreground toward the upper-right middle ground. Position the child in the upper-right quadrant, the koi across the lower foreground, and the bucket close to the right border. Keep generous warm neutral negative space above and around the book. Use shallow but coherent depth of field, crisp focus on the child and flooded pages, and softly receding page edges.
Lighting and color: Use a large diffused warm key from the upper left, soft ambient fill, and gentle shadows falling toward the lower right. Preserve controlled water highlights, readable separation between the blue dress and pale background, soft sheen on fish scales, and open detail in the aged paper. Use warm beige #C9C1B1 for the background, aged parchment tan #C8A66A with dark brown #6F5130 for page edges, muted aqua #6FAFB3 and pale blue-green #A7CFCE for the water, deep navy #1F4A73 for the dress, koi orange #F26A2E with warm ivory #F3E6D0, and bucket red #A72F2B.
Materials and finish: Render thick layered paper, brittle torn fibers, damp parchment, fine sand, translucent shallow water, glossy koi scales, matte painted metal, and soft woven dress fabric with physically plausible contact shadows, refraction, reflections, and wet-edge transitions. Aim for a handcrafted miniature sculpture photographed as a premium photorealistic 3D render, whimsical but grounded, clean, bright, detailed, and gallery-ready.
Text: Keep the book’s printed areas as dense, weathered, abstract typographic texture only. No readable words, letters, names, dates, logos, signatures, or watermarks.
Preserve exactly one open book, one child, one fishing net, one red bucket, three bright koi fish, the flooded-page illusion, the diagonal book placement, the warm neutral background, and the high-angle storybook composition. Avoid extra people, extra limbs, duplicated fish, warped anatomy, floating feet, oversized props, modern plastic objects, clutter, muddy shadows, blown highlights, fake HDR, cartoon styling, flat illustration, accidental text, logos, and watermarks. Output one high-resolution square image.