Novità: GPT Image 2.5 Flare e Sunburst. Provali gratis con 3 immagini alla registrazione, poi 8 crediti ciascuna.
Prova GPT Image 2.5
Idol elfo nella metropolitana coreana
Cinematico / Fotogramma

Idol elfo nella metropolitana coreana

Genera una scena cinematografica anime-realismo con un idol elfo accanto a un poster in metro. Personalizza nome del personaggio, colore dei capelli e titolo.

Prompt

A cinematic anime-realism scene inside a modern Korean subway station, featuring a life-size fantasy elf idol standing beside a massive illuminated character-design advertisement poster for [SERIO]. Composition & Camera: Vertical 3:4 composition, eye-level camera, realistic station perspective, shallow depth of field focusing on the foreground subject with blurred pedestrians in the background. Main Subject: A tall, elegant female elf standing on the left, facing forward with one hand gesturing toward the poster. She has very long flowing [orange] hair, bright blue eyes, long pointed ears, a delicate anime face, and a calm, slightly serious expression. She wears a luxurious ivory-white fantasy concert gown made of draped satin fabric with a high thigh slit, translucent shawl sleeves, orange ribbon accents, gold jewelry details, thin waist ornaments, anklets, and white strappy high-heel sandals. Anime styling blended seamlessly with photorealistic fabric textures and skin rendering. Setting & Lighting: Korean subway platform corridor with tiled walls, dark ceiling panels, metal columns, and glossy floor reflections under warm fluorescent indoor lighting. Overhead hang exactly two wayfinding signs: one large yellow sign with a left arrow, the number "3", and Korean text for exit; directly beneath it, a smaller dark sign with Korean transit transfer text. Poster: A huge backlit white advertising poster framed in black mounted on the right wall. At the upper left, large serif orange headline text reads "[SERIO CONCERT]" with smaller date text "2026.5.21" and brief Korean event notes beneath. The poster is organized into exactly nine visual sections: 1 large bust portrait of the elf, 1 full-body render, 1 Korean profile text block, 1 design-point section with 3 sketch callouts (ear jewelry, torso detail, leg/shoe detail), 1 expression section with exactly 8 small head portraits (arranged 4 on top, 4 on bottom), 1 fabric-flow sketch area, 1 dynamic-pose area with exactly 2 action poses, 1 key-scale section with exactly 2 crystal accessory or weapon sketches, and 1 makeup-detail area featuring exactly 2 close-up blue eye drawings. Constraints: Single coherent scene, not a collage. No extra main characters. Realistic microtext and clean typography.

Cosa puoi personalizzare

Prima di generare, sostituisci queste parti del prompt con il tuo soggetto o testo.

  • [SERIO]segnaposto
  • [orange]segnaposto
  • [SERIO CONCERT]segnaposto
  • “2026.5.21”testo tra virgolette

Dettagli del prompt

Modello
GPT IMAGE 2
Costo per immagine
1K: 43 crediti · 2K: 86 crediti · 4K: 143 crediti
Proporzioni
3:4
Aggiunto il

Altri modelli