Prompt

No image available.

This recipe has no reference image. Review the prompt details, then continue with the safe Image handoff when you are ready.

Kō

Kōda

@aimikoda

View on

82Likes6KViewsAug 24, 2026

Customize prompt

Fill the fields below before opening the matching model workspace.

themeRequiredstyleRequired

THEME = [Theme] STYLE = [Style] PALETTE = [Color Palette] ASPECT RATIO = [Aspect Ratio] DO: Create an artistic multi-scene collage built entirely around THEME. Derive subjects, environments, objects, characters and visual events naturally from THEME. Derive the emotional tone, atmosphere and energy naturally from THEME. Use people, animals, objects, landscapes, architecture, abstract forms or any combination of them when they meaningfully support THEME. Treat each scene as a distinct emotional or conceptual fragment from the same world. Build the composition like an expressive editorial art spread rather than a storyboard or regular panel grid. Freely vary the number, scale, proportions and placement of scenes. Combine large dominant images with narrow vertical fragments, wide horizontal crops, small detail scenes and partially cropped compositions. Stagger scene edges and shift their alignments to create an irregular visual rhythm. Let some scenes extend fully to the canvas edge while others terminate against neighboring imagery. Use occasional L-shaped, stepped or offset arrangements created naturally by neighboring scene proportions. Avoid repeating the same rectangle size or alignment across the composition. Keep scenes visually separate without visible frames or empty gaps between them. Let neighboring scenes meet through irregular crop relationships while keeping their environments independent. Do not blend separate locations into one continuous environment. Keep the main subjects of each scene visually contained within that scene. Use architecture, large color fields and negative space to shape the overall composition. Connect neighboring scenes through color relationships, visual rhythm and thematic echoes rather than physically merging their content. Choose environments and situations specifically suited to THEME. Include occasional intimate fragments such as hands, objects, windows, silhouettes, architectural details or empty spaces when they strengthen THEME. Give each scene a different visual scale, point of view and conceptual focus. Use expressive lighting, shape, perspective and negative space according to STYLE. Vary the emphasis of PALETTE from scene to scene while preserving overall harmony. Maintain consistency whenever subjects, characters, objects or locations recur. Keep the final composition visually clean, sophisticated and intentionally irregular. CAMERA: varied editorial framing, environmental wides, cropped medium views, isolated details, unusual negative-space compositions, overhead fragments and off-center perspectives chosen according to THEME RULES: Use STYLE as the authority for medium, linework, texture, shading, shape design and overall rendering language. Use PALETTE as the dominant color system across the image. Let THEME determine subjects, setting, mood, symbolism and visual narrative. Every scene must explore a different facet of THEME. The composition must feel assembled, layered and rhythmically irregular rather than divided into uniform boxes. Keep individual scenes independently readable while allowing their shapes and proportions to interact compositionally. Use asymmetry, scale contrast, staggered alignment and varied cropping to create the collage structure. Do not use visible panel outlines. Render every surface freshly, with continuous natural texture flow. Keep grain fine, stochastic, and evenly distributed across the frame. Give each repeated element its own unique micro-variation. Keep flat areas smooth with clean, unbroken tonal gradients. Resolve fine detail crisply at native resolution, as a first-generation render. NO: rigid grid, equal-sized panels, repeated rectangular boxes, visible panel borders, white gutters, margins, outer padding, blended environments, fused scenes, characters spanning multiple scenes, conventional comic-page layout, speech bubbles, captions, text, logos, watermarks

Output details

Media type: Image prompt
Task type: Text To Image

Model: GPT Image

Continue exploring

All Prompts

Prompt details: THEME = [Theme] STYLE = [Style] PALETTE = [Color Palette] ASPECT RATIO = [Aspect Ratio] DO: Crea

Reference preview

More like this