Original document
GPT Image 2 · en
Layered paper-cut illustration of a cozy chibi couple, rendered as a papercraft diorama with a handcrafted aesthetic. Technique: multiple stacked paper layers with soft drop shadows between each layer, creating 5 to 7 visible depth planes from foreground to background. Edges are smooth, rounded, and slightly beveled paper-cut edges, with subtle paper grain and fibrous texture visible on all surfaces. Soft, diffused inner shadows sit beneath each layer to suggest physical depth. Character design: chibi, cute, simplified proportions with a large round head and small body at roughly a 1:1.5 ratio; small dot eyes with a glossy highlight; soft circular rosy blush patches on the cheeks; an absent or minimal dot nose; a simple small curved smile; short, rounded, stubby limbs; and a clean, smooth silhouette with no sharp corners. Color palette: warm, cozy, pastel mood using soft cream, dusty rose, sage green, warm peach, sky blue, and honey yellow tones, kept at low-to-medium, muted and gentle saturation, against a warm off-white or soft gradient background. Lighting: soft ambient light from the top-front, with gentle white edge highlights on the top layers and warm light tan or beige shadow tones beneath the cut layers. Environment: simplified rounded leaf and floral shapes built as separate paper layers, curved horizon layers suggesting rolling hills, and tiny decorative paper-cut elements such as stars, small hearts, and dots. Overall mood: warm, whimsical, cozy, handmade, storybook. Render quality: ultra-detailed papercraft art with studio photography lighting and sharp focus on the layer edges.How to use
- Copy the complete original prompt, including any parameters.
- Replace the subject, text and composition with your own requirements.
- Use a compatible external tool; check its current access conditions and review the result.
Before you start
- Check whether a reference image is required. A compatible model label does not mean this example was generated with that model.
- Keep names, figures and visible text accurate; results can vary between model versions.
