Original document
GPT Image 2 · en
Based on the given [THEME], automatically generate a collectible epic narrative poster: a large, elegant silhouette of a person's profile serves as the outer contour, and within the silhouette, the most fitting complete worldview, iconic scenes, character relationships, symbolic motifs, key architecture, creatures, props, and atmosphere for that theme organically grow and fill the shape. The overall effect should not be an ordinary collage, but rather an elevated silhouette-fill narrative composite, evoking a double-exposure association, leaning more toward a fusion of cinematic poster and dreamy watercolor illustration style; soft aerial perspective, gentle misty transitions, paper grain texture, feathered edges and brush strokes, generous negative space, a restrained and premium layout, with a quiet, grand, sacred, nostalgic, poetic, and legendary feeling. The style, color palette, scenes, and materials should all automatically adapt to the theme, with every element strongly tied to the theme and instantly recognizable — avoid clutter, avoid hard collage effects, avoid templated backgrounds, and avoid cheap fantasy stock assets. Naturally incorporate the exclusive signature "[SIGNATURE_TEXT]" into the composition as part of the poster design, placed subtly but clearly — in the bottom-left corner, bottom-right corner, or near the title — in a style consistent with the overall layout, like the artist's signature or design seal of a collector's edition poster; the signature typeface should be refined, restrained, and premium, not too large, must not disrupt the main composition, and must not appear jarring or cheap.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.
