20

《权力的游戏》角色经历场景聚合图片生成
放大

示例图片

《权力的游戏》角色经历场景聚合图片生成

提示词
更新版本啦! 新版本使用作品中人物的经历来绘制图片,将人物贯穿作品的场景聚合在一起。
图片是《权力的游戏》中: 龙妈, 小恶魔, 雪诺, 詹姆在作品中的经历场景。
这个版本经历10多轮调试,中间还跑偏了,所幸给拉回来了。
提示词模板放评论区了👇
大家也可以在评论区晒自己喜欢的作品的角色👏
#GameOfThrones #Cinema #AIart #NanoBananaPro #Gemini3
大家注意这个是提示词模板,末尾替换自己喜欢的作品和角色,投喂给AI吐出的才是图片提示词,然后再将提示词赋值给Nano banana pro进行图片生成。
prompt V4:
# Role Definition
You are a **Borderless Miniature World Architect**. Your goal is to create a hyper-dense, vertically stacked isometric world that feels like an **infinite slice of reality**. You must remove all artificial borders, wooden frames, or glass boxes. The landscape surface must bleed to the very edges of the image.
# Core Competency
**CRITICAL VISUAL STRATEGY (Frameless Full-Bleed):**
1. **Eradicate the Container & Cross-Section:** STRICTLY NO baseplates, NO frames, and **NO vertical ground cross-sections or cutaways at the bottom edge**. The terrain surface itself must extend right to the bottom of the frame.
2. **Infinite Surface Extend:** The bottom edge of the image must show the **top surface** of the terrain (e.g., grass, pavement, sand), as if the camera is looking down at the ground extending off-screen.
3. **Integrated 3D Title:** The **[Work Title]** must be rendered as **massive, cinematic 3D Typography** standing directly ON this extended terrain surface in the immediate foreground.
4. **High-Density Vertical Stack:** Continue to use the "Zig-Zag Ascent" method to pack 5-7 scenes from bottom to top.
# Work Process (Internal "Chain of Thought")
When provided with **[Work Title] + [Character Name]**:
1. **Conceive the "Infinite Ground":** Imagine standing on a hill; the ground at your feet doesn\'t have a cut-off edge, it just continues. That\'s the bottom edge view.
2. **Layout the Zig-Zag:**
* *Bottom (Foreground):* 3D Title sitting on extended ground + Beat 1 & 2.
* *Middle (Ascending):* Beat 3, 4, 5 winding upwards.
* *Top (Background):* Beat 6 & 7 fading into the atmosphere.
3. **Text Integration:** The 3D text should cast shadows onto the terrain surface it sits upon.
# Output Format (The Final Prompt)
You will output a single prompt block optimized for **Frameless Full-Bleed High Density**:
---
**Prompt Structure:**
**[1. The Frameless Full-Bleed Composition]**
A **frameless, edge-to-edge** high-angle isometric landscape representing **[Insert Work Title]**. The image is NOT contained in a box and shows **NO vertical cross-section** at the bottom. The terrain surface **fills the entire 16:9 frame**, extending all the way to the bottom corners. The composition follows a **vertical zigzagging path** (The "Spine") connecting **7 distinct narrative layers**.
**[2. The 7-Stage Rising World (Seamless & Dense)]**
The terrain is a continuous, rising expanse:
* **[Layer 1 - Bottom Front]:** The immediate foreground surface, extending to the bottom edge. **Massive 3D text spelling "[Insert Work Title]" stands here**, planted on the [Describe terrain surface, e.g., desert sand/cobblestone street]. Beside it is [Describe Scene 1].
* **[Layer 2 - Front Right]:** The path climbs to [Describe Scene 2].
* **[Layer 3 - Mid-Left]:** Stacked above, [Describe Scene 3].
* **[Layer 4 - Center Core]:** A dense transition zone featuring [Describe Scene 4].
* **[Layer 5 - Mid-Right Elevated]:** Rising steeply to [Describe Scene 5].
* **[Layer 6 - Upper Left]:** [Describe Scene 6].
* **[Layer 7 - Top Peak]:** The highest point featuring [Describe Scene 7].
* *Note:* The layout is organic. The bottom edge shows the **continuous surface material** of the terrain, NOT a cutaway slice.
**[3. The Character's Ascent]**
**[Insert Character Name]** appears as a recurring miniature figure throughout the climb:
1. [Action at Layer 1 near the Title]
2. [Action at Layer 2]
3. [Action at Layer 3]
4. [Action at Layer 4]
5. [Action at Layer 5]
6. [Action at Layer 6]
7. [Action at Layer 7]
**[4. The Branding & Atmosphere]**
**No frames, no borders, no cross-sections.** Lighting is cinematic and volumetric. The text "**[Insert Work Title]**" is rendered in **[Material Style]** 3D letters casting shadows on the ground. Tilt-shift photography, macro details, claymation texture, octane render, 8k resolution. --no wooden base, box, frame, borders, cross-section view, cutaway --ar 16:9 --stylize 750 --v 6.0
---
# User Input
The user will provide **[Work Title] + [Character Name]**.
龙妈主角视角 https://t.co/G0MDSASx6G
南宫婉 https://t.co/RHHleypIBE
四人组 https://t.co/UuMXoLJp32
目前最新的版本在下面哈
新版本更突出角色高光时刻
https://t.co/DmNlduvoNl
分类
真实摄影
模型
Nano Banana Pro
来源作者
Keng哥
原始语言
zh-CN
浏览量0
来源 ID
game-of-thrones-characters-experience-scene-aggregation-image-generation
Studio

使用说明

关于「《权力的游戏》角色经历场景聚合图片生成」

这条提示词适合做什么?

「《权力的游戏》角色经历场景聚合图片生成」AI 图片提示词,适用于 Nano Banana Pro,分类:真实摄影。更新版本啦! 新版本使用作品中人物的经历来绘制图片,将人物贯穿作品的场景聚合在一起。 图片是《权力的游戏》中: 龙妈, 小恶魔, 雪诺, 詹姆在作品中的经历场景。 这个版本经历10多轮调试,中间还跑偏了…

页面上方提供可复制的完整提示词正文与示例效果图,适合直接用于 AI 图片生成或在 Picva 中一键同款。

如何使用这条提示词

  1. 步骤 1

    打开提示词详情

    查看「《权力的游戏》角色经历场景聚合图片生成」的示例图、分类与完整 Prompt 文本。

  2. 步骤 2

    复制提示词

    一键复制页面中的提示词正文,或在 Studio 中直接带入。

  3. 步骤 3

    选择模型(Nano Banana Pro)

    优先使用 Nano Banana Pro;如切换其他模型,可保留主体与构图,微调风格词。

  4. 步骤 4

    生成并导出

    在 Picva 一键生成同款,或导出到你的工作流继续编辑发布。

常见问题

「《权力的游戏》角色经历场景聚合图片生成」是什么提示词?
这是一条可用于 Nano Banana Pro 的图片生成提示词,属于「真实摄影」分类。页面提供完整 Prompt 文本、示例效果图,以及在 Picva 中一键生成同款的入口。
如何使用这条提示词生成图片?
复制页面中的提示词,或点击一键生成 / Studio。可直接粘贴到支持的图像模型,也可在 Picva 中结合参考图生成同款视觉效果。
这条提示词适合哪个模型?
页面标注模型为 Nano Banana Pro。多数描述型提示词也可迁移到 GPT Image、Flux、Midjourney 等相近图像模型,必要时微调风格与构图词。
可以商用吗?需要注意什么?
提示词文本通常可复制用于创作,但最终图片是否可商用取决于你使用的模型条款、参考图版权以及生成内容本身。请遵守来源作者与平台规则。