
Document original
GPT Image 2 · en
Scene:
Using the referenced host identity and the same thumbnail system, create a founder-debate spotlight cover. Preserve the left-heavy host crop, the right-side hook block, and the narrow top category strip. Keep the dark navy base, but switch the accent to vivid magenta and white with one split-circle or versus shape faintly behind the placeholder headline zone.
Subject:
The same recurring female host from the reference image appears unchanged in identity, hair, face, blazer, and lighting. She now points slightly toward the right-side hook area with a more skeptical expression, as if framing a debate between two public figures. The background should include only simplified silhouette medallions or abstract profile shapes, not full extra faces. The cover must read immediately as part of the same creator channel while signaling a debate or opinion segment.
Important details:
Identity preservation and template consistency are the main tasks. Keep the same overall camera distance and layout rhythm so all four covers look like one system when viewed side by side. The magenta accent should be decisive but limited. Use bold unreadable line blocks for the hook zone and clean graphic separators for a premium short-video feel. Do not let the supporting graphics compete with the host's face.
Use case:
Short Video Cover Pack series study 4/9, testing opinion-segment variation while maintaining one host-led cover system.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no extra realistic portrait collage, no different host styling, no overcrowded debate graphics, no muddy low-contrast layout.Mode d’emploi
- Copiez le prompt original complet, paramètres compris.
- Adaptez le sujet, le texte et la composition à vos besoins.
- Utilisez un outil externe compatible ; vérifiez l’accès et le résultat.
Avant de commencer
- Vérifiez si une image de référence est nécessaire. La compatibilité ne prouve pas le modèle utilisé pour cet exemple.
- Vérifiez noms, chiffres et textes ; les versions du modèle peuvent produire des résultats différents.





