AI Image Prompts · 绘画提示词库
出处与授权
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
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。
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
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 702 条
Pixar Style Cool Boy Illustration
comic-storyboardA 3D digital illustration prompt in the style of Pixar, featuring a character with specific accessories and a minimalist background.
展开提示词全文
A detailed 3D Pixar-style digital illustration of a {argument name="character" default="cool young boy with voluminous, wavy white hair and dark sunglasses"}. He is sitting on a small wooden chair in a relaxed, cross-legged pose against a stark black background. He is wearing a simple {argument name="outfit" default="light gray t-shirt, dark teal or green trousers, and chunky white sneakers"}. His right hand holds a lit cigarette, and his left hand holds a {argument name="prop" default="white Starbucks coffee cup"} with the full green mermaid logo facing forward. The lighting is soft and stylized, casting clean highlights on his hair and clothes. The perspective is centered on the character, with ample negative space above and around him.
Sci-Fi Punk Japanese Woman View
comic-storyboardA multi-view prompt for a woman in a sci-fi punk outfit, specifying a white color scheme with orange accents on a clean white background.
展开提示词全文
elegant {argument name="ethnicity" default="japanese woman"} in her {argument name="age" default="30s"} who is cool and sophisitcated in a {argument name="style" default="scifi punk"} outfit on on white background. front view, side view, back view in a whiteoutfit with orange design lines
Samurai War Tent Cinematic Shot
comic-storyboardA cinematic motion-focused prompt for a scene inside a samurai war tent during a sandstorm, designed for Nano Banana Pro.
展开提示词全文
Cinematic motion shot inside a {argument name="location" default="war tent"} during a {argument name="weather" default="sandstorm"}. The older samurai is turning his body away from the Shogun, mid-motion, slightly blurred. His movement feels reactive, not controlled. In the background, another enemy guard is approaching from the front. The Shogun is now partially out of frame, no longer the focus. Dust and light beams fill the scene. Dynamic composition, motion blur, cinematic realism.
The Painting That Refused To Be Finished
comic-storyboardA creative prompt for a surreal scene where a giant painted forest on a desert canvas begins to turn into a 3D living environment.
展开提示词全文
An enormous canvas — taller than a building — stands in the middle of a {argument name="environment" default="barren cracked desert"} at golden hour. The painting on it is almost complete: a {argument name="painted scene" default="hyperrealistic landscape of a lush green forest"}. But at the edges of the painted forest, the paint is actively peeling away and the trees are crawling off the canvas into the real desert — painted trees becoming three-dimensional, roots cracking into the dry earth, leaves casting real shadows. A {argument name="observer" default="tiny human figure"} stands at the base of the canvas looking up, dwarfed by its scale, unsure whether to step into the painted world or pull the trees back. The boundary between painted and real is the most detailed part — half flat brushstroke, half living texture. On the wooden frame of the canvas, Vertical 9:13 format. Ultra photorealistic, 8K, no text overlays, cinematic color grading.
Unraveling Figure Effect Comparison
comic-storyboardA surreal prompt for Nano Banana Pro and ChatGPT 1.5 that creates a photorealistic person whose arm is dissolving into thousands of colorful threads.
展开提示词全文
A standing male figure, photographed from the waist up, whose entire right arm and shoulder is slowly unraveling into {argument name="thread colors" default="thousands of red and black threads"} like a sweater being pulled apart the threads float and drift upward into darkness above. His face is {argument name="expression" default="stoic"}, looking down at his dissolving hand with quiet acceptance. Background is a {argument name="background" default="deep matte black"}. Left side of body is perfectly normal, photorealistic skin. The contrast between the real and the unraveling is the entire story. Studio lighting, single soft key light from the left. Thread textures ultra-detailed, each strand individually visible. On one of the longer drifting threads near the bottom, visible only on close inspection. Vertical 9:13 format.
Hyper-realistic Cyborg Side Profile
comic-storyboardA high-detail prompt for generating a half-human, half-cyborg portrait with a mix of digital painting and architectural sketch styles.
展开提示词全文
A hyper-realistic side profile of a {argument name="subject" default="man"}, half-human half-cyborg. Left side of the face is deconstructed to show complex {argument name="armor color" default="red"} armor plates, gears, and wires. Sharp, melancholic gaze. Style: detailed digital painting mixed with {argument name="background style" default="white background architectural sketches"}.
High-contrast lighting, 8k resolution, cinematic. ar 4:5
AI-generated 4-panel manga: The Vanishing Magic Ball
comic-storyboardA title for an AI-generated 4-panel manga created using Nano Banana Pro, titled 'The Vanishing Magic Ball,' featuring Carp and Camp.
展开提示词全文
AI image generation 4-panel manga Title The Vanishing Magic Ball Carp and Camp I wonder if the names are okay... Nishimura-san... It's just a personal hobby...
Cinematic diptych of teenagers on a rooftop
comic-storyboardA prompt for generating a cinematic diptych composition featuring a teenage boy and girl in Japanese school uniforms on a rooftop, focusing on close-up portraits and a nostalgic, slice-of-life mood.
展开提示词全文
Cinematic diptych composition of a teenage boy and girl in Japanese school uniforms on a rooftop. Top frame: close-up portrait of the girl with short bob hair, soft natural smile, looking at camera, subtle camera focus frame overlay. Bottom frame: close-up of the boy holding a DSLR camera, looking through the lens, capturing the moment. Soft natural daylight, clear sky, calm atmosphere. Minimalist composition, strong negative space, clean framing. Slice-of-life storytelling, nostalgic mood
3x3 Storyboard Grid of a Young East Asian Woman
comic-storyboardA highly detailed prompt for generating a 3x3 storyboard grid featuring a young East Asian woman in consistent clothing and setting, specifying nine different poses and compositions for commercial product photography.
展开提示词全文
3×3 storyboard grid, nine panels, same subject and outfit, consistent teal studio background, soft diffused lighting, clean editorial lifestyle photography young East Asian woman, natural makeup, black hair tied back, wearing white fitted top, blue/cream color-block cardigan with red strap, blue jeans, sneakers, pastel blue over-ear wireless headphones panel sequence (left→right, top→bottom): 1 standing adjusting headphones, front view 2 close-up portrait touching earcup 3 side profile sitting on tall stool 4 lying on stomach with cassette tapes + retro player 5 extreme close-up headphones + face detail 6 tilted head smiling portrait 7 low-angle crouch reaching headphones toward camera 8 forced perspective product shot, headphones close to lens 9 lying on back relaxed pose style: minimal, pastel tones, soft shadows, sharp focus, commercial product photography, consistent framing, balanced grid layout, mix of wide/medium/close shots
Manga page prompt for Nano Banana Pro
comic-storyboardA prompt for Nano Banana Pro to generate a single manga page in a 'manga style' (漫画風) featuring two sisters discussing something important that one cannot remember.
展开提示词全文
Create one page of a manga featuring these two people. White background with black lines.
Manga theme: {argument name="manga theme" default="An older sister who absolutely cannot remember something important, and her younger sister who cannot guess what it is"}
The manga panels progress from the top right to the bottom left.
Nano Banana Pro 4-Panel Manga Prompt: Absolute Zero Ignore
comic-storyboardA prompt used to generate a 4-panel manga titled 'Absolute Zero Ignore,' depicting extreme indifference and a 'cannon of disinterest' being fired.
展开提示词全文
AI Image Generation 4-Panel Manga Title Absolute Zero Ignore Super ignore A cannon of disinterest explodes If the goal was to cause a flame war Serves you right! I guess? lol
Nano Banana Pro 4-Panel Manga Prompt: The Ultimate Choice
comic-storyboardA prompt used to generate a 4-panel manga titled 'The Ultimate Choice (Religiously),' focusing on the spiritual cost (karma) of receiving 'all wealth' rather than its physical value.
展开提示词全文
AI Image Generation 4-Panel Manga Title The Ultimate Choice (Religiously) Presenting the 'spiritual cost (karma)' of receiving 'all wealth' rather than its physical value.
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
更新记录
官方源监视中内容有更新(细节请以官方文档为准)
内容有更新(细节请以官方文档为准)
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。
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。
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。
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。
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。
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。