


Create one vertically stacked before-and-after comparison using the supplied image as the only source. LAYOUT:Exactly two equal-height panels: original image on top, transformed image on bottom. Both panels have the source image's aspect ratio and identical width. The complete canvas is twice the source image's height. Join the panels directly at a clean horizontal seam. No thick divider, extra panel, decorative frame, labels, or added text. TOP PANEL — ORIGINAL:Place the supplied image unchanged in the upper half. Preserve its full composition, original colors, lighting, details, existing text, and crop. Do not redraw, retouch, stylize, or apply texture to this panel. BOTTOM PANEL — OUTSIDER ART:Redraw the same scene as a flat, densely textured naïve folk-art print. Preserve the main subject, recognizable silhouettes, relative positions, important objects, and overall composition. Do not invent additional subjects or props. Flatten the ground into map-like color zones, with people, trees, buildings, and objects shown in simple frontal or side elevation. Remove converging perspective, photographic depth, directional lighting, and cast shadows.Use warm paper white, warm ink black, and two to four subdued colors derived from the source. Choose one existing object or connected detail as the only saturated accent. Give each large zone one consistent handmade texture: stitch marks, irregular window grids, or dense foliage marks. Keep the texture dense, rhythmic, and slightly imperfect. Where people appear, use small, solid, faceless figures with filled clothing and limbs of visible width. Render every surface as opaque printed ink on matte paper, with no photographic remnants.Keep all illustration, paper texture, and color changes strictly inside the lower panel. No gradients, glossy vector finish, watercolor, stick figures, cute cartoon proportions, invented lettering, or watermark.
This image was generated with GPT Image 2.5 by OpenAI. The prompt on this page is the exact text used to create it.
The example was generated with GPT Image 2.5 using: 3:4 · Abstract · Vintage · Sketch. Reusing these settings gives the closest match.
Copy the prompt as-is first, then replace only the subject while keeping the style, lighting, and composition keywords. Generate a few variations — outputs differ between runs.
Yes. Prompts work best as starting points — adjust descriptors, aspect ratio, or style tags to make the output your own.