Image prompt Posters & Typography Customizable
Wuxia Movie Poster With Breaking-the-Frame Hero
A vertical Eastern movie poster built from three stacked spring-garden photos, with the central figure bursting past the black dividers. The cross-frame effect gives the flat collage real depth and a cinematic lift.
- Model
- GPT Image 2
- Type
- 4 images
- Aspect ratio
- 2:3
- Shared
- 28 Jul 2026
- Likes on X
- 37
CharacterSignature textLabel text
Image prompt
3,092 charactersUsing [Character] as the subject, create a 2:3 vertical live-action COS Eastern wuxia movie poster. The image uses an Editorial Magazine three-layer horizontal collage composition. The whole is made of three horizontal photos of equal width filling the canvas horizontally, with nearly equal heights and thin black dividing lines between layers. The first layer is a distant view of a spring Eastern garden environment, including cherry blossoms, a lake surface, old trees, pavilions, and empty sky. The figure is positioned slightly right, holding the character's weapon, establishing the full environment and world. The second layer is a horizontal close-up of the figure, positioned on the right, highlighting the face, makeup, eye color, hairstyle, iconic accessories, and costume details; the left retains the garden, flowering branches, lake surface, and natural depth of field. The third layer is the hero-visual figure layer of the entire poster, shot with an ultra-wide-angle at close range. The third layer is not an ordinary photo confined to the bottom frame, but the topmost layer above all collage photos. The figure occupies about 80% of the poster; the head, body, long hair, sleeves, ribbons, and weapon extend upward from the third layer, clearly breaking through its boundary, crossing over and covering the second layer, and partially entering the first layer. The figure must break the black dividing lines, with body and weapon covering partial areas of the two prior layers, forming a clear Out-of-Frame, Breaking the Grid, Magazine Cutout, and Hero Image Overlay naked-eye 3D cross-frame effect. The first and second layers exist as background photos; the third-layer hero figure image floats atop the entire collage layout and cannot be confined inside the third layer's rectangle. The third layer uses a strong-perspective ultra-wide-angle composition; the weapon, arm, sleeve, or ribbon near the lens is clearly enlarged, the face stays sharp, the levels are clearly defined, creating a spatial depth that bursts out of the frame. Preserve [Character]'s original hairstyle, hair color, eye color, iconic accessories, weapon, thematic color scheme, and core design elements, transforming the costume into a high-fidelity live-action COS version using real embroidery, brocade, light gauze, metal ornaments, and layered fabric. The overall style is neo-Chinese wuxia, Eastern fantasy, spring garden, commercial COS photography, high brightness low saturation, clear airy feel, soft film colors, cinematic lighting, real skin texture, and real costume materials. Add a simple stylized "V" and cat-element graphic logo in the top-left corner; add a small "voxCAT" signature in the bottom-right, with vox lowercase and CAT uppercase; subtly incorporate VOXCAT woven labels, metal nameplates, or hidden-pattern details into parts of the costume, belt, or weapon. The image is clean and sharp, figure proportions natural, hand structure accurate, costume details stable, the three-layer collage relationship clear, and the hero image's cross-frame effect strong.
How to use it
Start with a photo of your own character or a clear reference image, since the look depends on a real person and their costume. It was shared as a GPT Image 2 prompt, so keep the three-layer structure and the hero figure breaking the grid, which carry the effect. Swap the character, weapon and costume colours to suit your story, and keep the cherry blossoms and lake for the spring atmosphere. Works well as a poster mockup or a bold profile banner.
Featured inMovie poster AI prompts
Shared publicly on X on 28 Jul 2026. All credit for this prompt and its result goes to its creator. Is this yours? Ask for an update or removal.

















