
Original document
GPT Image 2 · en
Scene:
A square short-video thumbnail for a true-crime channel, designed for maximum mobile-size readability. The composition uses a dark blue-black background, one flashlight beam, one close human face on the left, and a stark empty hook zone on the right. A narrow red top strip signals the category. The overall layout should feel clean, urgent, and highly clickable without becoming a messy poster.
Subject:
An investigative female host in a dark coat leans into frame holding a flashlight upward, her expression tense but controlled. Behind her, a crime-scene wall and evidence board appear only as large abstract shapes, not detailed clutter. The right side contains bold non-readable headline bars and one simple caution-tape diagonal. The frame must read instantly at thumbnail scale.
Important details:
Keep the face large, the lighting dramatic, and the design extremely simple. Use red only as a category accent so the face remains the focal point. The hook zone should feel intentional and strong even with no real words. This is a channel cover system, not a cinematic key art poster.
Use case:
Short Video Cover Pack single study 5/9, focused on true-crime thumbnail clarity and bold hook-zone placement.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no cluttered evidence collage, no multiple faces competing, no muddy darkness, no tiny props everywhere.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.





