
Original document
GPT Image 2 · en
Scene:
A tall 9:16 romance cover inside a hospital vending-machine alcove at 3 a.m., with dim fluorescent spill, stainless steel reflections, soft green walls, and one warm vending-machine glow illuminating tired faces. The poster should be optimized for phone screens: a compelling upper silhouette, a center emotional beat, and a lower quiet gradient for imagined title and episode-chip zones. Keep the palette cool mint, navy, pale skin tones, and one gentle warm amber highlight.
Subject:
Create a night-shift romance short-drama cover featuring an overworked resident doctor holding instant noodles and a sleep-deprived paramedic leaning against the vending machine with a blood-pressure cuff hanging from one wrist. Their bodies angle toward each other in an almost-kiss distance, but the mood is exhaustion and mutual dependence rather than glamour. Use eye lines, fatigue detail, and the narrow space between them to make the hook instantly legible on a mobile device.
Important details:
The image should feel platform-ready and addictive without becoming cheesy. Preserve realistic scrubs, ID-lanyard shapes that stay unreadable, vending-machine light, and hospital textures, but keep the frame uncluttered. The design logic should imply where a title, subtitle, and small platform badge would live while never actually rendering them. Emotional clarity at thumbnail scale matters more than background complexity.
Use case:
Vertical Drama Poster System single study 6/9, testing a night-shift romance hook in a narrow mobile-cover format.
Constraints:
No readable text, no watermark, no logos, preserve exact 9:16 aspect ratio, no medical emergency chaos, no giant heart graphics, no broad comedy, no crowded hallway extras, no app screenshot frame, no candy-colored rom-com palette, no broken hands or faces.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.





