# 超现实交互场景

Canonical: https://prompt.genedai.me/prompts/gpt4o-085/
Category: 超现实与创意
Model (source label): GPT-4o / gpt-image-1（合集分类）
Requires reference image: false
Snapshot: 2026-10-19

把现实场景和超现实动作结合，明确谁与谁互动以及接触点，减少空间关系混乱。 替换提示词中的[主体1]和[主体2]为具体的主体描述，例如"一个女孩"和"一朵玫瑰"。

## Prompt

```
A pencil sketch of [Subject 1] interacting with [Subject 2], where [Subject 2] is rendered as a realistic, full-color object, creating a surreal contrast against the hand-drawn style of [Subject 1] and the background
```

## Source collection Chinese translation

一幅铅笔素描画，描绘了 [Subject 1] 与 [Subject 2] 互动的场景，其中 [Subject 2] 以逼真的全彩风格呈现，与 [Subject 1] 及背景的手绘素描风格形成超现实的对比。

## Attribution and evidence

- Prompt author: @umesh_ai https://x.com/umesh_ai
- Original source: https://x.com/umesh_ai/status/1917444534239191544
- Collection: https://github.com/jamez-bondos/awesome-gpt4o-images
- License: CC BY 4.0（仓库与案例声明；原作者权利仍须保留）
- License URL: https://github.com/jamez-bondos/awesome-gpt4o-images/blob/main/cases/85/ATTRIBUTION.yml
- Image source: https://raw.githubusercontent.com/jamez-bondos/awesome-gpt4o-images/main/cases/85/case.png
- Image provenance: 维护者依据对应 Prompt 用 Sora 重新生成的示例；不是原帖原图。
- Date: 2025-05-24 (ATTRIBUTION 的复现日期；不是原帖发布日期)
- Evidence: 合集 GitHub stars（不是单图点赞）; historical dataset values, not current popularity or a reproduction test.

Third-party rights and upstream exceptions apply. Preserve attribution and disclose changes.