
Oryginalny dokument
GPT Image 2 · en
Scene:
A square steak hero shot framed from a slightly elevated three-quarter angle, with a large white porcelain plate resting on polished black marble under controlled overhead light and a narrow warm side kicker. The lighting should reveal lacquered surfaces, gentle steam, and plate curvature while keeping the background clean and dark. The image must feel like a high-budget restaurant advertising visual rather than casual dining photography.
Subject:
Feature three neatly sliced pieces of medium-rare wagyu beef arranged in a staggered line over a glossy dark glaze. Add a small smoked onion puree swoosh, a few charred pearl onions, tiny mustard greens, and one precise arc of reduced jus. The beef should show realistic marbling, moist cut faces, and a restrained amount of rendered fat sheen. Use a very small portion, plated like a tasting course, with one faint wisp of smoke or vapor rising near the meat to imply heat and service timing.
Important details:
Wagyu texture must be convincing, with fine marbling and correct pink center rather than raw or overcooked extremes. Keep the glaze reflective but controlled, and avoid barbecue styling. The plate should read as European fine dining: minimal components, strong negative space, and exact sauce geometry. Every garnish must support the meat and not turn the plate into a busy still life.
Use case:
Michelin Plate Hero Shot single study 3/8, focused on premium wagyu treatment with lacquered surfaces and restrained drama.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no steak knife, no hands, no wooden board, no oversized ribeye slab, no casual grill marks, no messy grease, no extra side dishes, no cartoon smoke clouds.Jak używać
- Skopiuj cały oryginalny prompt wraz z parametrami.
- Dostosuj temat, tekst i kompozycję do swoich potrzeb.
- Użyj zgodnego narzędzia zewnętrznego; sprawdź warunki dostępu i wynik.
Przed rozpoczęciem
- Sprawdź, czy potrzebny jest obraz referencyjny. Zgodność nie potwierdza modelu użytego do przykładu.
- Sprawdź nazwy, liczby i napisy; wyniki zależą od wersji modelu.





