← 返回首页
技能

AI Image Prompts · 绘画提示词库

github · 1.0.4 · 下载量 1,092 · 热度分 90.0

出处与授权

作者: YouMind-OpenLab
许可证: MIT 许可证 (原文)
版权声明: Copyright (c) 2026 YouMind-OpenLab

本站只收录项目元数据与简介,不对作者正文做转载;完整内容请到上方出处查看。商用前请自行核对原仓库的 LICENSE 与 NOTICE。

展开许可证全文
MIT License

Copyright (c) 2026 YouMind-OpenLab

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

LICENSE · 2026-09-26 23:14:14

产品介绍

Recommend curated prompts from a 10,000+ real-world image generation prompt library. Works with ANY AI image model — Nano Banana Pro, Nano Banana 2, Seedream 5.0, GPT Image 1.5, Midjourney, DALL-E 3, Flux, Stable Diffusion, and more. Use this skill when users want to: - Find proven image generation prompts (any model) - Get prompt inspiration for portraits, products, social media, posters, etc. - Create illustrations for articles, videos, podcasts, or marketing content - Browse categorized prompt templates with sample images - Translate and understand prompt techniques

资源信息

作者:YouMind-OpenLab
分类:图文音视频多模态生成
语言:en
版本:1.0.4
认证方式: 无需认证
许可证:MIT
是否付费: 官方标注免费
发布方:YouMind

许可证信息摘自官方源,可能过时或缺失。商用前请以原仓库的 LICENSE 文件为准。

标签
绘画提示词文生图营销素材

安装方式

npx skills i YouMind-OpenLab/ai-image-prompts-skill

功能说明

  • 按用途分成 11 类,不是随机堆砌
  • 全文已入库,站内直接检索、可复制
  • 语义检索(描述需求即可)
  • 内容改写模式(贴正文生成定制提示词)
  • 数据每日同步两次
  • MIT 许可,无需 API Key
适用场景
找某类图片的现成提示词给文章/视频/播客配插画做海报 / 封面 / 社媒图翻译并理解提示词写法

提示词库 · 2.3万

数据源:社区提示词库(MIT 许可,每日同步)

共 15141 条提示词,按 11 个用途归类;同一条可归入多个用途,因此分类条数相加会大于去重后的总数。

命中 224 条

Cinematic 12-Panel Timeline of a Subject's Epic Day

youtube-thumbnail

This is a detailed image generation prompt designed to create a 12-panel (4x3) cinematic timeline. It captures an epic, chaotic, or hilarious day in the life of a specified subject, complete with timestamps, punchy captions, and richly detailed scenes. The overall visual UI, backdrop, and panel frames are themed to match the subject's universe, such as arcane scrolls for wizards or chrome HUD for cyborgs.

展开提示词全文
"In the Day of {argument name="subject" default="[SUBJECT]"}": Design a 12-panel (4x3) cinematic timeline capturing an epic, chaotic, or hilarious day in the life of {argument name="subject" default="[SUBJECT]"}. Add this title in a style that matches the subject to the top of the image. Each panel must be [PHOTOREALISTIC], large, and show:
a timestamp (e.g. 06:00, 08:00...)

a short, punchy caption in quotes (e.g. "Potion Mishap", "Midnight Bounty Run")

a richly detailed, high-energy scene tailored to that moment of their day.
The overall backdrop, panel frames, textures, and visual UI elements must match the subject’s universe — e.g., {argument name="theme example" default="leather + studs for rockstars"}, arcane scrolls for wizards, chrome HUD for cyborgs, jungle vines for pirates, etc.
Panels are arranged in a grid or storyboard, with rugged or themed dividers. Final panel (06:00 next day) shows collapse, rest, or post-chaos resolution.

Nano Banana Pro Thumbnail Generation: Street & Pop Style

youtube-thumbnail

This is a prompt used for generating a thumbnail image for 'Nano Banana Pro' with a 'Street & Pop' theme. The creator suggests using a street-style aesthetic to make the thumbnail stand out and incorporating a horse as the main character to appeal to Japanese users interested in the Chinese zodiac.

展开提示词全文
Nano Banana Pro Thumbnail Generation

Theme: {argument name="Theme" default="Street & Pop"}

Using a "{argument name="Style" default="Street Style"}" aesthetic, which is rarely seen online, can make it relatively stand out. By making the main character a {argument name="Main Character" default="Horse"}, it also attracts the interest of Japanese people who care about the Chinese zodiac.

Moody Cinematic Portrait Prompt for AI Image Generation

youtube-thumbnail

This is an image generation prompt designed to create a moody, dramatic cinematic portrait with intense side lighting and a thriller-like gaze. The prompt specifies a high-contrast, high-detail aesthetic suitable for generating striking visual art.

展开提示词全文
A moody cinematic portrait of a {argument name="subject" default="man"}, dramatic side lighting, high contrast, intense gaze, deep shadows, high detail, thriller aesthetic, {argument name="style" default="photorealistic"}, 8k

Cinematic Mountain Portrait with Floating Text

youtube-thumbnail

A prompt for generating a realistic cinematic portrait of a confident young man in a misty mountain landscape. The image features a muted teal color grading, shallow depth of field, and the glowing text “2026” subtly floating in the sky, creating a calm, futuristic mood.

展开提示词全文
A realistic cinematic portrait of a confident young man standing in a misty mountain landscape, wearing a black t-shirt, blue cargo pants, sunglasses, and a backpack, hands in pockets, soft foggy atmosphere, muted {argument name="color grading" default="teal"} color grading, shallow depth of field, minimal aesthetic background with distant layered hills, the glowing text “{argument name="text content" default="2026"}” subtly floating in the sky, natural soft light, ultra-realistic photography, 85mm lens, high detail, calm futuristic mood.

Nano Banana Pro Composition Techniques Prompts (3 Examples)

youtube-thumbnail

A tutorial-style tweet explaining three advanced composition techniques for Nano Banana Pro: 1) using 'Text Anchors' to force negative space for thumbnails/posters, 2) using 'Physical Camera Work Specification' (like wide-angle, low-angle) for dynamic scenes, and 3) using 'Infographic Grid' for organized layouts like process diagrams. It includes executable prompts for each technique.

展开提示词全文
[Role]
You are a professional graphic designer. Create a YouTube thumbnail.

[Composition Instructions]
- Split-screen composition, dividing the screen into left and right halves.
- Left side: A Japanese woman with a surprised expression (photorealistic style). Background is a dark office.
- Right side: A vibrant yellow background. Place the text “{argument name="text" default="AI Revolution"}” in bold Gothic font in the center.
- Ensure sufficient negative space around the text to maximize visibility.

[Style]
A scene from an animated movie, cyberpunk

[Camera and Composition Settings]
- Lens: Use a wide-angle lens (24mm equivalent) to emphasize perspective.
- Angle: Extreme low angle (worm's-eye view). The character is viewed looking up from just above the ground.
- Placement: Place the character's left hand directly in front of the camera to exaggerate the sense of distance.
- Background: High-rise buildings sharply converge toward a vanishing point.

[Role]
You are a manual creator. Create {argument name="subject" default="how to make Okonomiyaki"}.

[Format]
Flat design infographic, 16:9

[Layout Instructions]
- A “3-step” process diagram flowing from left to right.
- Step 1 (Left): Magnifying glass icon (Investigation).
- Step 2 (Center): Gear icon (Processing).
- Step 3 (Right): Light bulb icon (Solution).
- Connect each element with arrows and align them at equal intervals.
- Background is white, icons are unified in blue and orange key colors.

Jay Chou Blessing Image Integration Prompt

youtube-thumbnail

A Nano Banana Pro prompt designed for image-to-image generation, specifically to seamlessly integrate text (a blessing from Jay Chou) from a source image (Image 1) into a new main image (Image 2). The prompt emphasizes maintaining the original image's composition while ensuring the blessing text is small, readable, and harmonizes with the lighting and visual elements of the final cover.

展开提示词全文
Redraw a high-quality, refined, and aesthetically pleasing cover image. Seamlessly and **completely integrate all Jay Chou–related information from Image 1 into Image 2**, placing it at a **small, delicate scale** and in a **well-balanced, appropriate position**, As if the blessing from Jay Chou were a **naturally integral part of Image 2’s composition**, with the blessing cleverly echoing the core visual elements of the main image and sharing a unified, consistent lighting design. The final cover should convey a **warm, beautiful atmosphere of blessing**. 
**Important:** Ensure the blessing text is **clear and highly readable**. When necessary, use tasteful techniques such as contrasting/inverted text colors, Gaussian-blurred backgrounds, or other elegant visual enhancements to improve legibility while maintaining visual harmony.

Nano Banana Pro Image Generation Prompt for MBTI Character Visualization

youtube-thumbnail

An extensive image generation prompt for Nano Banana Pro, designed to visualize an MBTI personality type as an anime-style character. It takes the MBTI diagnosis results (type and reason) from a previous LLM analysis and uses them to customize a base character profile (anime girl with brown hair and red eyes). The prompt includes detailed instructions for UI layout (9:16 vertical image, text placement), character layering rules based on the 16 MBTI types (e.g., ENTJ gets a military-style jacket and a crown), and specific color themes for each MBTI group.

展开提示词全文
## 1. Input Variables (Rewrite this)
Generate an illustration based on the following information.

User Information:
target_account: "{argument name="Account Name" default="@schnapoon"}"
mbti_type: {argument name="MBTI Type" default="ENTJ (Commander)"}
reason: "{argument name="Diagnosis Reason" default="Posts structurally share how to utilize AI tools and actively engage with users. Efficient prompt creation and creative idea proposals stand out, showing a leadership-oriented thought pattern."}"

Character Setting (base_character_profile):
Appearance: Anime-style girl. Brown medium hair, red eyes, slightly pointed teeth (giza-ba).

## 2. Generation Instructions
A. Overall Image Composition (UI Layout)
Generate a vertical 9:16 image. The background color is clean off-white (#F5F5F7).
The screen is composed of the following elements from top to bottom:

1. Header Text:
   Place the text `target_account` + "'s diagnosis result is" at the top of the screen.
   Color: Dark gray (#333333), Font size: Medium, Alignment: Center.

2. Main Character Illustration:
   Important (Consistency): Maintain the "physical features (face, hair, body type)" and "basic clothing (innerwear, bottoms, etc.)" of the `base_character_profile` without alteration.
   Changes (Layering): Express the individuality of each type by "layering" the specified outerwear, mantle, and decorations from "B. MBTI Style Definitions" below over the basic clothing.
   Presentation: Draw specified effects and floating items around the character.
   Style: High-quality anime coloring.

3. MBTI Type Label:
   Place the `mbti_type` text largely directly below the illustration.
   Color: Theme color of each type definition.
   Font: Bold for high visibility.

4. Diagnosis Reason Text Box:
   Place a white, rounded-corner box (card-like design) at the bottom of the screen.
   Write the `reason` text in Japanese inside it.
   Text color: Gray (#555555), Left-aligned, readable size.

B. MBTI Style Definitions (16 Types Layering Rules)
Apply the style matching the input `mbti_type` and add the following items over the basic clothing:

[Analysts - Purple/Dark Blue/Deep Red]
   INTJ (Architect):
       Layering: Black long coat or a chic jacket with a standing collar.
       Items: Chess pieces, hologram of blueprints.
   INTP (Logician):
       Layering: Lab coat (slightly disheveled), or a loose hoodie.
       Items: Flasks, floating mathematical formulas, gadget parts.
   ENTJ (Commander):
       Layering: Military-style Napoleon jacket, or a heavy mantle with a red lining.
       Items: Baton, crown-like decoration.
   ENTP (Debater):
       Layering: Unconventional design blazer, or a magician-like cape.
       Items: Microphone, smirking expression, jack-in-the-box effect.

[Diplomats - Green/Gold/Pastel]
   INFJ (Advocate):
       Layering: Hooded robe, or a mystical shawl.
       Items: Lantern, staff, mystical light particles.
   INFP (Mediator):
       Layering: Oversized knit cardigan, or a forest-girl style poncho.
       Items: Flower crown, butterflies, books.
   ENFJ (Protagonist):
       Layering: Hero-like mantle, shining brooch or sash on the chest.
       Items: Holy sword (light-shaped), halo-like effect.
   ENFP (Campaigner):
       Layering: Colorful patchwork vest, or a flashy scarf.
       Items: Balloons, confetti, party goods.

[Sentinels - Blue/White/Teal]
   ISTJ (Logistician):
       Layering: Neat navy blazer, armband.
       Items: Clipboard, glasses, neatly arranged documents.
   ISFJ (Defender):
       Layering: Clean apron, or medical coat/nurse uniform-style outerwear.
       Items: First aid kit, warm drink (tea, etc.).
   ESTJ (Executive):
       Layering: Formal suit jacket,

Youtube Thumbnail Concept: The New Weapon

youtube-thumbnail

A prompt for generating a YouTube thumbnail concept split 50/50 between a grayscale 'War' side and a high-saturation 'Fusion' side, with a central cutout of Trump acting as the divider, featuring large yellow text reading 'THE NEW WEAPON'.

展开提示词全文
Youtube thumbnail Concept:
Background (Split 50/50):

Left Side (War): Grayscale or desaturated sepia tone. Shows a chaotic background of smoke or tanks.
Right Side (Fusion): High-saturation, deep blues and glowing purples. Shows a futuristic tokamak reactor core.
Subject: A large, shoulder up cutout of Trump placed dead center, acting as the divider between the two worlds. He should look confident/determined, facing slightly toward the "Fusion" side.
Text (Top, full width):

"{argument name="main text" default="THE NEW WEAPON"}" (in large, blocky #FFFF00 Yellow font).
Note: Add a heavy black drop shadow or a dark gradient vignette at the very top to ensure the yellow text is readable against the glowing background."

Book Promotion: AI Utilization for Diagrams Prompt Guide

youtube-thumbnail

A promotion for a book detailing AI utilization techniques for creating diagrams, including a guide to image generation prompts for Nano Banana Pro to create 'diagrams that convey information in 2 seconds.'

展开提示词全文
[Nano Banana Pro Compatible] AI Utilization Techniques to Turn Anything into a Diagram: Convert to a 'Diagram that Conveys in 2 Seconds'! Complete Guide to Image Generation Prompts for Business

Nano Banana Pro Prompt for a Game Strategy Guide Style

youtube-thumbnail

A user shares a prompt designed to generate images in the style of a game strategy guide using Nano Banana Pro, encouraging others to try it out easily.

展开提示词全文
Nano BananaPro prompt in the style of a game strategy guide.

Nano Banana Pro Prompt for Generating 'Viral Diagrams'

youtube-thumbnail

An announcement about an updated guide that includes a 'secret prompt' for generating 'viral diagrams' instantly using Gemini 'Nano Banana Pro,' along with prompts for creating diagram bots and four-panel comics.

展开提示词全文
Secret prompt revealed for instantly generating "viral diagrams" with Gemini "Nano Banana Pro"

・ Diagram creation Bot
・ Four-panel comic creation

Existential Cinematic Solitude Film Still

youtube-thumbnail

A prompt for creating a cinematic film still drenched in emotional weight, inspired by 1970s existential cinema (Vilmos Zsigmond). It focuses on profound solitude, cold blue hour lighting, rain-streaked glass distortions, and a muted, desaturated palette, using technical details like analog grain and ray-traced reflections to create crushing atmospheric pressure.

展开提示词全文
[{argument name="subject" default="Subject"}], profound solitude captured in cinema, hyper-real film still drenched in emotional weight, cold blue hour lighting with dying warmth at the edges, rain-streaked glass creating liquid distortions, shallow depth of field emphasizing loneliness, muted desaturated palette with isolated color accents, slow-motion implied stillness, fine analog grain reminiscent of 1970s existential cinema, lens flares bleeding into darkness, ray-traced reflections in pooling water, breath visible in cold air, negative space dominating the frame, crushing atmospheric pressure, intimate yet distant camera positioning, Vilmos Zsigmond-inspired melancholic cinematography, the weight of unspoken grief.

提示词模型无关,可直接用于 Midjourney / DALL·E / Flux / Stable Diffusion 等任意文生图模型。

依赖说明

- Node.js 18+
- 无需任何 API Key(数据全部公开)
- 提示词已完整同步到本地与数据库,站内检索不依赖第三方

适用平台

claude-codecursorcodexgemini-cliopenclaw

更新记录

官方源监视中
内容更新 2026-10-19

官方源有内容更新(首次纳入监控)

信息边界说明

本页信息摘自官方文档,仅作技术信息呈现,不代表本站的判断或推荐。价格、配额、可用性以官方最新说明为准。你离开本站后的任何行为(申请 Key、付费、部署、商用),本站不做任何承诺与担保。

同类推荐

视频短视频图片音乐音频配音口播数字人全能创作

官方标注收费

通过同一个AI创作工具完成AI图片、视频、音乐与语音创作,查找公开社交数据,编辑视觉成果,并集中查看和管理生成资产。Beatra 全能创作支持AI图片生成(文生图)、AI视频生成(图生视频)和AI视频编辑,通过AI音乐生成把想法做成歌曲或纯音乐,也可用AI语音生成(文字转语音)、多语言旁白和AI语音克隆完成跨媒介内容;还能查找抖音数据、小红书数据,以及 YouTube、TikTok、Instagram 与 X 上的公开帖子、笔记、评论、账号和趋势,把社交媒体数据用于竞品调研和创作。它是一款面向AI内容创作与多模态内容创作的AI创作套件,用起来就是一个把图片、视频、音乐与语音收在同一处的AI创作助手,适用于社交内容、产品展示、营销素材、课程、播客、短片和跨媒介项目;还能在不同媒介间复用原始素材、已有成果和查到的公开资料,查看制作进度、返回文件、实际用量和积分记录。

AI数字人口播与虚拟主播视频

官方标注收费

用一张人像和一段短脚本或语音,制作有明确讲述内容的数字人口播视频。可使用现成录音,也可以先选择音色,把脚本生成旁白,让画面中的人物随声音完成稳定、自然的口播。适合 AI 主播、虚拟主播视频、数字人讲解、让照片说话、照片说话视频、产品讲解视频、数字人培训视频、公告和社交口播,并在交付后检查人物一致性、声音清晰度、口型同步和动作稳定性。

罗斯蒙德可视化图形生成

官方未说明

【罗斯蒙德可视化图形生成】南京罗斯蒙德基于已交付项目沉淀的可视化图形生成场景:根据业务描述或系统需求,自动生成可编辑的架构图、流程图、ER 图、时序图、状态图与 C4 图。采用「统一图模型」一次建模、多格式输出:结构化文本(Mermaid / PlantUML,可嵌入 Markdown、进 Git 版本管理)、可编辑图形文件(drawio / Excalidraw,可拖拽修改节点与样式)、静态图(SVG 自渲染,PNG 给转换命令),并输出图形说明(图的含义 + 关键节点解释)。内置样式系统统一配色与命名,内置布局引擎自动排布,校验脚本拦截孤立节点、悬挂连线与语义缺漏。触发词:画架构图、画流程图、画 ER 图、画时序图、画状态图、C4 图、系统架构图、部署架构图、数据流程图、实体关系图、生成 drawio、生成 Excalidraw、导出 mermaid、导出 PlantUML、导出 SVG、导出 PNG、可编辑的图、图转可编辑、架构图可编辑、流程图二次修改、画图、出图、画个图。

架构图一键生成

官方标注收费需申请 Key

使用 ContextWeave 生成或修改架构图、流程图、思维导图和复杂信息图。适用于需要从代码、文件或自然语言中提取结构与关系并实际产出 CW、SVG、HTML、可编辑 PPTX 或原生 VSDX 的请求;未特别说明时,PPTX 与 Visio 使用原生形状和连接器导出。不适用于统计图表、手绘插画或像素级排版。

架构图一键生成

官方标注收费需申请 Key

使用 ContextWeave 生成或修改架构图、流程图、思维导图和复杂信息图。适用于需要从代码、文件或自然语言中提取结构与关系并实际产出 CW、SVG、HTML、可编辑 PPTX 或原生 VSDX 的请求;未特别说明时,PPTX 与 Visio 使用原生形状和连接器导出。不适用于统计图表、手绘插画或像素级排版。

海报设计工作室

免费增值需申请 Key

将一段活动描述、一张产品照片或品牌参考图,转化为抓人眼球的活动海报、宣传 banner 或社交媒体配图。这款 AI海报设计与制作工具生成图片丰富的海报,搭配清晰视觉层级、按品类匹配的风格和预留文字区域,覆盖活动、音乐节、电影、产品发布和促销场景。你可以把它当作 AI宣传单制作工具来制作促销宣传单,当作活动海报设计工具,当作电影海报生成器,当作音乐节海报设计工具,当作产品发布会海报的 banner设计 AI,或当作平面设计海报工作室。从活动主题出发,用品牌风格参考图变换产品或场景照片,或在已认可草稿上做定向精修,完成可印刷或可分享的成品。

什么是技能

技能(Skill)是封装好的单项能力,比如文字识别、文档摘要、关键词提取。开发者把调用逻辑和提示词打包,使用者像用插件一样加载即可,无需了解底层模型。启枢技能 收录大量技能,可直接试跑或编入方案。