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 1099 条
Emotional Watercolor Grief Art
othersAn expressive watercolor painting capturing sorrow and vulnerability through indigo tones and tear-streak details.
展开提示词全文
A delicate, expressive watercolor painting on {argument name="paper texture" default="textured white cold-press paper"} capturing the close-up profile and facial features of a person in emotional turmoil. The artwork features a striking color palette of {argument name="eye colors" default="deep teal, indigo, and violet"} around the expressive eyes, transitioning down the nose into warm shades of peach and yellow, and finished with vibrant, glossy crimson lips. A thin, prominent tear streak drips down from the lower lid in a fluid indigo wash, blending minimalist negative space with moody, bleeding pigments to evoke a profound sense of sorrow and emotional vulnerability.
Mother of Dragons RAW Photography Style
othersA high-detail prompt simulating a RAW photograph of a fantasy queen character, focusing on intricate textures and deep focal depth.
展开提示词全文
{ "visual_details": { "subject": "Woman resembling Daenerys Targaryen, detailed portrait", "hair": "Platinum blonde, intricate braided hairstyle on top, loose wavy tendrils framing the face, sharp and clear texture", "expression": "Neutral, focused expression looking slightly off-camera", "attire": "Pale grey pleated gown with deep plunging neckline, gold filigree metal waist belt, matching gold filigree collar. Fabric texture and metal details are highly defined.", "pose": "Leaning forward on a dark wood table surface, arms extended downward. Sharp focus extends across her hands and the table texture.", "background": "Dark interior setting with visible, sharp architectural details, including dark wood paneling, a stone wall texture, and defined window frames. No depth of field blur.", "lighting": "Warm, natural light illuminating the scene, with distinct golden backlighting originating from the table surface and background elements, remaining sharp and defined rather than soft." }, "raw_photo_technical_details": { "format": "Unprocessed CR3 RAW file data", "resolution": "High-resolution 4K UHD (3840x2160)", "camera_settings": { "shot_on": "Canon EOS R5", "lens": "RF 50mm f/1.8 STM", "aperture": "f/11 (Deep depth of field, entire scene in focus)", "shutter_speed": "1/160 sec", "iso": "640 (Natural digital grain visible)", "white_balance": "Daylight" }, "texture": "Realistic skin texture, visible fabric weave, defined metal details, natural digital noise/grain throughout the image", "focus": "Pan-focal, sharp detail from foreground (subject) to background architectural elements." }, "modified_prompt": "A sharp, high-resolution RAW photograph (4K UHD) of a woman resembling Daenerys Targaryen, detailed with intricate platinum blonde braids and wavy tendrils. She wears a pale grey pleated gown with a plunging neckline, a detailed gold filigree belt, and collar. She is leaning forward on a dark wood table with her arms extended. The scene is illuminated by warm, natural light with defined golden highlights. The entire image exhibits deep focus (pan-focal), with sharp, clear detail extending to the dark wood and stone architectural elements in the background, which have no blur. Natural digital grain, realistic skin texture, and sharp fabric details are visible, characteristic of an unprocessed RAW photo shot at f/11 aperture." }
Driftplane Exposure System Prompt
othersA meta system prompt used to instruct LLMs on how to generate stylized image prompts for Nano Banana Pro.
展开提示词全文
DRIFTPLANE EXPOSURE Rewrite the user's text, image, or both as one prompt for GPT Image Gen V2 or Nano Banana Pro. Use GPT unless Nano is named. Write in the user's language, preserve exact text, and keep requested or source ratio. Use three systems. Plane Field fixes structure through blue or cyan planes, silhouettes, horizons, reflection, shadow, and negative space. Exposure Drift records elapsed motion through smear, ghosting, blur, edge loss, color drag, and displaced light. Signal Trace marks the event with one warm or green accent where motion, contact, or reflection leaves evidence. Do not make a blurry night photo or clean blue poster. Determine subject, plane anchor, moving force, drift direction, trace source, receiver, consequence, contact sites, and quiet region. Preserve subject, count, identity, action, setting, clothing, objects, colors, text, and references. For edits, preserve unrequested content. Use style references only for blur, color drag, blue geometry, shadow, reflection, and negative space. Do not copy people, cups, dogs, houses, boats, trees, or grid. Plane Field needs one blue family, crisp boundaries, scale, and plausible material. Reflections belong only to water, glass, floor, or metal. Shadows follow the receiving plane. Exposure Drift needs cause and direction. It may come from walking, vehicle motion, handheld shake, wind, water, smoke, light sweep, heat, crowd flow, or subject movement. Keep one drift axis. Do not blur every object. Stable anchors stay readable. Signal Trace may be red, orange, amber, green, or warm white. It must attach to a source, reflection, object, moving body, or contact mark. No random neon, full-frame wash, or duplicated limbs. Use one sharp plane, one drift field, one trace accent, one focal contact, and clear negative space. Remove repetition, praise, mood labels, prestige terms, fake technical detail, generic defects, and alternatives. Use concrete nouns and active verbs. No em dash or en dash. GPT: return 500 to 800 words in five paragraphs covering scene, subjects, plane field, drift, trace, contacts, composition, preservation, text, and failures. NANO: return valid JSON only, 1000 to 1800 tokens, using "aspect_ratio", "references", "scene", "subjects", "plane_field", "exposure_drift", "signal_trace", "contact_sites", "composition", "surface_and_optics", "text", "avoid". Omit unused fields. No metadata, IDs, weights, comments. Return only the finished prompt.
Textured Impasto Oil Portrait
othersA prompt for a captivating oil painting portrait featuring thick impasto brushstrokes, palette knife textures, and a moody, elegant atmosphere.
展开提示词全文
A captivating oil painting portrait of a woman with {argument name="hair style" default="long, dark, wavy hair"} and expressive {argument name="eye color" default="green-hazel"} eyes, wearing a {argument name="clothing" default="textured light-colored shirt"}. The art style features thick, visible impasto brushstrokes and heavy palette knife textures, creating a dimensional, tactile feel. The background is a dynamic mix of deep blues, ochres, and golden-browns with expressive, abstract strokes. The lighting highlights the warm skin tones, rosy cheeks, and soft pink lips, conveying a moody yet elegant atmosphere.
Quiet Library and Traditional Style Portrait
othersA clean prompt for generating portraits set in a quiet library, focusing on traditional styles and a love for books.
展开提示词全文
A {argument name="setting" default="quiet library"} and a {argument name="style" default="bright red traditional look"}. A love for books & style.
Song-based Image Generation (Himawari no Yakusoku)
othersA creative prompt that uses a Japanese song title to generate an associated visual mood and scene.
展开提示词全文
{argument name="song title" default="Himawari no Yakusoku"} ({argument name="artist" default="Motohiro Hata"})
Zentangle Squirrel Line Drawing
othersAn intricate prompt for a black and white zentangle-style squirrel drawing, perfect for adult coloring pages.
展开提示词全文
A black and white line drawing of a {argument name="subject" default="squirrel"} is depicted in profile, facing left. The {argument name="subject" default="squirrel"} is decorated with various patterns and designs, resembling a {argument name="art style" default="zentangle or adult coloring page"}. Its body is filled with floral motifs, scales, swirls, and geometric shapes. Its tail is large and curled, also adorned with patterns. The background is white, emphasizing the detailed linework of the {argument name="subject" default="squirrel"}.
Algae Candies Still Life
othersA simple, creative prompt generating a whimsical and humorous still-life image of a specific subject at a nostalgic location.
展开提示词全文
image of a {argument name="subject" default="bowl of algae-flavored candies"} at {argument name="location" default="Grandma's house"}
Stonepunk Overgrown Temple Courtyard
othersAn artistic stonepunk-style image featuring a jungle-overgrown temple courtyard with intricate stonework.
展开提示词全文
{argument name="subject" default="A jungle-overgrown temple courtyard"} in an artistic {argument name="style" default="stonepunk"} style.
Man with Smiley Mask Portrait
othersA medium close-up portrait of a bearded man holding a weathered yellow smiley face mask.
展开提示词全文
A medium close-up, eye-level portrait of a {argument name="subject" default="man"} shown in a three-quarter profile facing right, holding a {argument name="accessory" default="distinctive accessory"}. He has dark messy hair, paired with a dense, sharply groomed full beard. He is wearing thick-rimmed, dark acetate sunglasses, a {argument name="outfit" default="tailored matte black blazer"}, and a deep plum-colored crewneck t-shirt. In his left hand, he holds up an oval, bright yellow mask featuring a simple black smiley face line, circular eye holes backed with black mesh, and realistic, weathered dark scuff marks across the surface. The lighting is soft and natural, emphasizing the textures of his hair and the contours of his profile against a smoothly blurred, neutral-toned outdoor background.
Cinematic Korean River Portrait
othersA serene, highly detailed portrait of a woman in a mountain valley near a crystal-clear river.
展开提示词全文
A highly detailed cinematic portrait of a beautiful {argument name="subject" default="Korean girl"} standing near a {argument name="location" default="crystal-clear river in a peaceful mountain valley"}. She has soft natural makeup, long silky black hair flowing in the wind, and is wearing a {argument name="outfit" default="light elegant summer dress"}. Surroundings are lush green mountains, misty atmosphere, golden sunlight reflecting on the water, ultra realistic, 8K, shallow depth of field, dreamy and serene mood, soft cinematic lighting, nature photography style.
Detailed Embroidery Textile Art
othersA style-transformation prompt that turns any uploaded image into intricate embroidery textile art with visible thread textures and stitches.
展开提示词全文
Recreate upload image into a detailed {argument name="art type" default="embroidery textile art"}, needlepoint style, {argument name="texture" default="intricate thread texture"}, visible stitches, yarn colors, cotton fabric backing texture, tactile and cozy feel, artisan craftsmanship.
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)
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。