APIPod logoAPIPod
不对称现代剪贴簿拼贴

不对称现代剪贴簿拼贴

不对称的现代剪贴簿风照片拼贴:完全保留参考图女子的面部,排版自由随性。

2026年8月18日4 次浏览

提示词

GPT Image 2

GPT Image 2 prompt: Asymmetrical modern scrapbook style photo collage featuring a young woman using the exact face from the reference image, preserving identity, proportions, and natural skin texture across all frames. Multiple cutout photos of the same woman in playful poses (jumping, peace sign, winking, hands on cheeks, cheerful expressions), arranged in a dynamic layered composition with polaroid frames, rounded cuts, torn paper edges, sticker-style frames, and transparent overlays. Hairstyle variations include long wavy hair, ponytail with thin bangs and baby hair, low bun with straight bangs, and fishtail braid with soft bangs, all clean and well-styled. Korean makeup look with dewy skin, soft glow, gradient lips, subtle blush, light shimmer eyes, clean eyeliner, natural lashes. Include cute pastel 3D chibi characters inspired by cell-like emotions interacting around the collage, along with hand-drawn doodles such as hearts, stars, sparkles, speech bubbles, and scribbles. Add scrapbook elements like pastel washi tape, memo notes, stickers, paper textures, and layered decorations. Outfit is casual stylish (oversized sweater or crop hoodie with high-waist jeans or skirt and sneakers) in soft pastel tones (pink, mint, lilac, baby blue). Background varies slightly per frame (aesthetic bedroom, cute café, or pastel textured backgrounds). Photography style includes close-ups and medium shots, slightly tilted angles, shallow depth of field (f/1.8–f/2.8), creamy bokeh, soft natural lighting with warm tones, ultra-sharp details, crisp edges, no blur or noise. Clean balanced composition in vertical 9:16 format. Add a small handwritten rose gold watermark on the left side reading ‘cr: 3v_ayra’ with a tiny butterfly. Negative prompt: blurry, low resolution, pixelated, distorted face, inconsistent identity, bad anatomy, messy hair, harsh shadows, cluttered layout, oversaturated colors, text errors, large watermark.

如何为 GPT Image 2 改写这条提示词

  • 1替换主体和场景时,保留光线与镜头语言(如黄金时刻轮廓光、85mm f/1.4)——这些细节决定写实质感。
  • 2保留构图线索(取景、背景景深、色彩基调),把风格关键词换成你想要的美学方向。
  • 3明确写出不想要的元素(无水印、无文字),并在模型支持时指定输出比例。

相关提示词

Botanical Scientific Illustration
图像

Botanical Scientific Illustration

A botanical scientific illustration of a blooming magnolia branch in the style of 18th-century naturalist art. Precise detail showing flower petals, stamens, leaves with visible veins, and a cross-section of the fruit. S

GPT Image 2
botanicalscientificillustration
Blog Article Hero Layout
图像

Blog Article Hero Layout

A blog article hero section with a full-width cover image of a mountain landscape at sunrise. Overlay gradient from transparent to dark at the bottom. White text overlay: category tag 'ADVENTURE', large serif headline 'I

GPT Image 2
uibloghero
Fashion Poster Layer Separation
图像

Fashion Poster Layer Separation

Separate a fashion campaign poster into model, garment, typography, background, and decorative design layers.

Seedream 5.0 Pro高级
layer separationfashion
Single-Prompt 16-Panel Storyboard
图像

Single-Prompt 16-Panel Storyboard

Generate a numbered 16-panel storyboard from one prompt, with consistent characters and coherent camera logic across the sequence.

Seedream 5.0 Pro进阶
storyboardsequence
Trading Terminal Interface
图像

Trading Terminal Interface

Generate a full trading terminal UI — candlestick charts, order book, bid/ask spread, volume, and timestamps — from one line.

Seedream 5.0 Pro入门
interfaceui
Pet Photo to Mecha Transformation
图像

Pet Photo to Mecha Transformation

Turn a plain pet photo into a detailed mecha version with a single short instruction, keeping the original pose and character.

Seedream 5.0 Pro入门
image inputtransformation

本提示词库中的所有内容均采集自公开互联网,仅供学习交流参考。 如有侵权,请告知我们,我们将第一时间删除。