
原文
GPT Image 2 · en
Scene:
A square illustrated portrait composition surrounded by sixteen small swatch-like source cues arranged as subtle border tiles, not a rigid reference grid. Warm paper texture, soft studio light, and refined editorial illustration quality.
Subject:
A single coherent portrait of an imagined traveler built from sixteen visual cues: round glasses, teal coat, embroidered scarf, canal-blue shadows, brick bridge texture, lantern warmth, compass motif, bicycle handlebar curve, cat-shaped pin, moss-green stair color, wooden window geometry, watercolor sky, ink cloud edge, food-stall red accent, film grain, and cream paper tone. The border swatches hint at the inputs while the central portrait is the fused result.
Important details:
The central character must feel unified, not like a patchwork costume. The sixteen cues should be discoverable through color, material, shape, and accessories. Use clean hierarchy, gentle texture, and clear silhouette. No readable labels are needed.
Use case:
Independent sixteen-reference fusion single image, showing many source cues merging into one character identity without external refs.
Constraints:
No readable text, no logos, no watermark, no real person likeness, no cluttered collage, no extra faces, no photorealism, preserve exact 1:1 aspect ratio.使い方
- パラメータを含む元のプロンプト全体をコピーします。
- 被写体、文字、構図を自分の要件に置き換えます。
- 対応する外部ツールで使用し、利用条件と結果を確認します。
開始前の確認
- 参考画像の要否を確認してください。互換モデルの表示は、そのモデルでの生成実績を意味しません。
- 名前、数値、画像内の文字を確認します。モデルの版で結果は変わります。





