
Documento original
GPT Image 2 · en
Scene:
Render the same square hero composition with a smartphone front and center, subtle layered depth, and studio-grade reflections that match the reference screen set. The image should look like a polished product-marketing tile focused on one feedback moment in the learning journey. Keep the square frame balanced, calm, and production-ready so the asset can serve gallery review, research comparison, and real educational delivery at the same time.
Subject:
Show a practice screen with picture choices, gesture-friendly controls, celebratory but controlled feedback icons, and a gentle correction state that still feels positive. Avoid text entirely and communicate meaning through color, motion cues, and a very obvious answer-feedback hierarchy. The subject matter should feel intentionally selected for a coherent curriculum family rather than like a random decorative assortment.
Important details:
Hold the reference system tightly: identical module shapes, mascot styling, palette anchors, and UI spacing. The scene should prove that the app can handle instruction and response with clarity, not just onboarding charm. Preserve the same mobile device framing, card radius, mascot style, color system, icon weight, and spacing rhythm from the reference screen set. Use variation only in the teaching content, supporting props, and scene-specific mood while preserving the core visual system exactly.
Use case:
Learning App Onboarding and Lesson UI series study 3/4, practice feedback screen.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no readable UI copy, no app-store branding, no logos, no generic gaming clutter, no impossible device anatomy, no notification spam.Cómo usar
- Copia el prompt original completo, incluidos los parámetros.
- Adapta el tema, el texto y la composición a tus necesidades.
- Usa una herramienta externa compatible; comprueba el acceso y el resultado.
Antes de empezar
- Comprueba si necesita una imagen de referencia. La compatibilidad no demuestra qué modelo generó el ejemplo.
- Revisa nombres, cifras y textos; el resultado puede variar entre versiones del modelo.





