GPT Image 2.5 Sunburst プロンプト - 柔らかな自然光のグレープフルーツ厨房ポートレート
- アスペクト比
- 9:16
- 画質
- 1K
プロンプト
A young East Asian woman with long dark brown hair tied back in a loose bun,
loose strands falling over her face and neck. She wears a white off-the-shoulder
top with a smocked chest panel and ruffled trim on the armholes, a delicate
heart-shaped pendant necklace, and a thin bracelet on her left wrist. She is
seated at a wooden kitchen counter, leaning slightly forward, hands resting
near the counter edge, looking directly at the camera with a subtle smile.
In the immediate foreground, on a wooden cutting board, is a cut half of a red
grapefruit; beside her is a clear glass bowl containing red apples and oranges.
A clear glass vase or pitcher sits on the left counter.
The background is a bright kitchen: large white-framed windows with green
foliage visible outside, white cabinets, a sink, and glass jars with dried
food items on the counter.
Shot at eye level, medium shot framed from the upper thighs up, subject placed
center-left. Estimated 50-85mm portrait lens with shallow depth of field:
her face is sharp while the background windows and the foreground fruit are
softly blurred. Soft diffused natural daylight from the windows, frontal and
slightly from her right, low contrast, gentle highlights on skin and fabric,
neutral-to-slightly-cool color temperature. High-key white and beige palette
with red and orange accents from the fruit and green through the windows.
Soft, dreamy, clean lifestyle aesthetic.
Preserve exactly: eye-level medium framing, white off-shoulder smocked top
with ruffled trim, heart-shaped necklace, cut grapefruit on cutting board,
glass bowl of apples and oranges, windowed bright kitchen background, soft
diffused natural light.
Do not add or change: no extra people, no extra props, no text or watermarks,
no heavy skin smoothing beyond natural look, no artificial studio lights,
no oversaturated colors.
この画像プロンプトを使う
- 01
プロンプトから始める
プロンプトをコピーするか、生成ボタンで開きます。
- 02
自分のアイデアを加える
被写体や背景、細部を調整し、必要に応じて参考素材を追加します。
- 03
生成前に確認する
モデルと出力設定を確認してください。生成結果は毎回異なる場合があります。







































