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 224 条
Candid Urban Street Photography Prompt (JSON Structure)
youtube-thumbnailA structured JSON prompt for Nano Banana Pro to generate a highly realistic, candid urban street photograph. It describes a young woman walking on a city sidewalk, holding a coffee cup, wearing layered clothing (leather jacket, sweater vest), and specifies camera settings (selfie-style, 28mm wide-angle) to achieve a cozy, cinematic street mood with earthy tones.
展开提示词全文
{
"image_generation": {
"subject": {
"description": "young woman walking alone on a city sidewalk",
"pose": "holding a takeaway coffee cup in one hand and a small paper bag in the other, arm slightly extended as if filming herself",
"expression": "calm, thoughtful, slightly introspective",
"gaze": "looking downward"
},
"appearance": {
"hair": {
"color": "dark brown",
"style": "messy bun with loose strands"
},
"face": {
"skin_tone": "fair with natural complexion",
"makeup": "minimal, natural makeup"
},
"accessories": [
"silver over-ear headphones"
]
},
"outfit": {
"top": "white collared shirt layered with beige sweater vest",
"outerwear": "{argument name="jacket color" default="brown"} leather jacket",
"bottom": "blue denim jeans",
"footwear": "casual sneakers"
},
"environment": {
"setting": "urban residential street",
"background_elements": [
"red brick apartment buildings",
"black iron railings",
"sidewalk scaffolding",
"city pathway"
],
"weather": "overcast"
},
"props": {
"drink": "white takeaway coffee cup with lid",
"bag": "small brown paper bag"
},
"lighting": {
"type": "natural daylight",
"direction": "overhead ambient light",
"quality": "soft and diffused"
},
"camera": {
"angle": "slightly high angle, selfie-style",
"framing": "medium shot",
"lens": "28mm wide-angle",
"depth_of_field": "moderate, background slightly blurred"
},
"aesthetic": {
"style": "urban lifestyle photography",
"vibe": "cozy, candid, cinematic street mood",
"color_palette": "earthy tones, browns, muted reds, neutral grays"
},
"quality": {
"realism": "highly realistic",
"resolution": "high resolution",
"look": "editorial street-style photograph"
}
}
}
Hyper-Realistic Candid Night Race Photo with iPhone Flash Aesthetic
youtube-thumbnailA highly detailed prompt for generating a hyper-realistic, chaotic, and energetic night-time photo simulating an unedited iPhone rear-camera flash aesthetic. It specifies a subject at a racing circuit, harsh head-on flash lighting creating abrupt shadow falloff, wide-angle distortion, and visible noise/grain, aiming for a spontaneous, unposed look identical to a reference image.
展开提示词全文
Subject: Person in the image Create a 3:4 hyper-realistic candid night-time event photo in an unedited iPhone rear-camera flash look. Shot handheld at chest-to-face height, close to the Subject, with a wide-angle feel (≈24–26mm equivalent), slight edge distortion, and deep depth-of-field keeping the Subject, nearby crowd, and racetrack background readable. Vertical medium-close framing from waist to slightly above the head. The Subject stands near a railing at a racing circuit, body angled slightly sideways, one arm bent casually near the railing, the other relaxed. Framing feels rushed and instinctive—slightly off-center, with background elements overlapping and partially cropped. Direct on-camera flash fires head-on, strongly illuminating the Subject and creating harsh highlights on skin, hair, and racing jacket, with abrupt shadow falloff behind the shoulders. Stadium lights, LED screens, and signage appear in the background as blown highlights and soft bokeh points. Low-level noise and grain remain visible in darker crowd and structure areas. Phone-style exposure prioritizes the Subject, lifting highlights and compressing shadows. Mild computational sharpening appears along hair strands, jacket seams, printed logos, and facial contours. The Subject’s expression is alert and slightly distracted, mouth gently parted, gaze directed off-frame as if reacting to live action nearby. The moment feels mid-event and unposed. The Subject wears a white-and-dark racing-style jacket with visible motorsport branding and patches over a fitted inner top, showing structured fabric, subtle sheen, and natural folds. A thin necklace catches small flash highlights at the collarbone, with jeans and a belt visible at the waist. The environment is a crowded night racing circuit with railings, spectators, sponsor signage, LED screens, lighting rigs, and structural beams. The background feels busy, noisy, and energetic, adding visual clutter and motion. The image should feel like a spontaneous iPhone flash photo taken during a live racing event—chaotic, energetic, and human. No cinematic lighting, no editorial posing, no beauty retouching. Do not change the Subject’s facial features or proportions. The Subject must look 1000% identical to the reference image.
High-End Cover Visual Generation Prompt
youtube-thumbnailA simple, effective prompt template for generating a high-end cover or feature visual. It specifies a centered composition, generous negative space, premium studio lighting, and a clean background to ensure clarity at both thumbnail and full scale.
展开提示词全文
High-end cover visual with a single dominant subject. Centered composition, generous negative space, soft shadows beneath the subject, premium studio lighting with subtle highlights, clean background with gentle gradient, designed to read clearly at thumbnail size and full scale.
Creating a Thumbnail Image with Text Overlay
youtube-thumbnailA simple prompt used with Nano Banana Pro's 'Image Editing' feature to create a thumbnail image based on a selected source image, adding a specific text overlay.
展开提示词全文
Create a thumbnail image using the text "{argument name="text" default="Update Complete!"}"
Ultra-Realistic Portrait of Female Bass Guitarist Performing Live
youtube-thumbnailA detailed prompt for generating an ultra-realistic, vertical portrait of a young female musician playing a bass guitar live on an outdoor stage. It specifies the subject's appearance, attire (acid-wash crop top, wide-leg jeans), instrument details, and technical parameters for high-fidelity, cinematic photography.
展开提示词全文
{
"image_generation_prompt": {
"subject": {
"main": "young female musician bass guitarist performing live on an outdoor stage",
"appearance": "long dark wavy layered hair blowing slightly in wind, fair skin, intense focused facial expression, natural makeup, silver hoop earrings",
"pose": "standing confidently holding a bass guitar, fingers plucking strings, looking slightly off-camera to the left"
},
"attire": {
"top": "sleeveless dark grey acid-wash graphic crop top with vintage band aesthetic",
"bottoms": "loose fit wide-leg blue denim carpenter jeans with large pockets",
"accessories": "black leather belt with double row silver grommets, wide black guitar strap"
},
"instrument": {
"type": "modern electric bass guitar",
"details": "deep blue burst finish with visible wood grain texture (quilted maple top), black hardware, long guitar neck, intricate fretboard details"
},
"environment": {
"setting": "daytime outdoor concert stage, open air festival vibe",
"background_elements": "blurred stage truss structures, black bass amplifier stack with metal grille, microphone stand on the left, overcast bright sky"
},
"technical_style": {
"quality": "ultra realistic, 8k uhd, hdr, masterwork, sharp focus, high fidelity textures",
"lighting": "natural diffuse daylight, soft shadows, cinematic lighting",
"camera": "telephoto lens, depth of field, bokeh background, portrait photography"
}
},
"parameters": {
"aspect_ratio": "9:16",
"orientation": "vertical long",
"resolution": "highest possible"
}
}
High-Quality Photo of Man Flying Above Clouds
youtube-thumbnailA prompt for generating a realistic, high-quality, wide-angle photo of a young man flying horizontally above a sea of white clouds, emphasizing a sense of freedom and joy through cinematic lighting and clear blue sky.
展开提示词全文
A realistic, high-quality photo of a happy young man flying horizontally in the sky above a sea of fluffy white clouds. He has his arms outstretched like wings and legs slightly bent. He is wearing a sage green hoodie, beige cargo pants, black sneakers, a black baseball cap, and sunglasses. The sky is clear blue with bright sunlight, creating a sense of freedom and joy. Cinematic lighting, 8k resolution, wide-angle shot.
Antarctica Penguin Adventure (2 Variations)
youtube-thumbnailA prompt offering two variations for generating photorealistic, cinematic images set in Antarctica. The first is a wide-angle selfie with a man surrounded by Emperor penguins and glaciers. The second is a detailed scene of a man kneeling to shake hands/flipper with a penguin, emphasizing scale and adventure.
展开提示词全文
Prompt Image 1:- A wide-angle selfie photograph in Antarctica showing a man sitting on the snow, surrounded by a large colony of Emperor penguins. The man has short dark hair, a full beard, and is wearing red-lensed metal-frame sunglasses, a dark blue unbuttoned denim jacket over a white crew-neck t-shirt, dark blue cargo pants, and a brown canvas backpack. He is leaning slightly back, holding an orange smartphone in his right hand to take the selfie, with his left hand near the backpack strap. Around him, dozens of adult Emperor penguins are standing and walking on the snow, some close to the camera and others stretching into the distance. The background features massive towering glaciers and ice mountains with white cliffs and blue ice formations, deep snow with visible footprints, and a clear bright blue sky. Bright natural daylight creates crisp shadows, with sharp focus on the man, penguins, and the vast icy landscape, captured from a slightly low perspective that emphasizes scale and isolation. Prompt Image 2:- A highly detailed, cinematic wide-angle photograph in Antarctica: a handsome young man with short dark neatly styled hair, a full trimmed beard, and a confident expression is kneeling on one knee in the deep snow, gently shaking hands / holding flipper with a large adult Emperor penguin standing upright in front of him. The man is wearing red-tinted metal-frame aviator sunglasses, a dark blue unbuttoned denim jacket over a white crew-neck t-shirt, dark blue cargo pants, brown canvas backpack with straps over both shoulders, white Nike sneakers, and a wristwatch. He is looking at the penguin with a slight smile, his right hand clasping the penguin's flipper. The Emperor penguin is life-size, black head with yellow-orange neck patches, looking directly at the man. Surrounding them is a vast icy Antarctic landscape with massive towering glaciers, snow-covered mountains, blue ice formations, and distant icebergs under a clear bright blue sky. Crisp natural daylight, sharp shadows, high detail on textures of snow, denim fabric, penguin feathers, and facial features. Slightly low camera angle to emphasize scale and adventure, photorealistic style, sharp focus throughout, vibrant yet natural colors & heartwarming composition.
Urban Night Portrait with 100% Facial Accuracy
youtube-thumbnailA detailed JSON prompt for generating a high-impact cinematic urban night portrait. It explicitly requires 100% facial accuracy matching a reference image, focusing on street style, strong reflections on wet asphalt, and cinematic color grading.
展开提示词全文
{
"image_type": "Urban Night Portrait",
"prompt": "High-impact cinematic depiction that matches the reference image with 100% facial accuracy. Featuring a young woman with long, wavy brown hair and a gentle smile. She is dressed in an oversized orange graphic t-shirt, a distressed denim mini-skirt, and clean white sneakers, complemented by a small brown crossbody bag. The subject is leaning casually against a metal pole on a wet urban street at night. The environment emphasizes a relaxed, modern city atmosphere with dramatic street lighting, neon accents, and strong reflections on the wet asphalt. Captured using a high-quality portrait lens with shallow depth of field, cinematic framing, and blurred city lights creating soft bokeh in the background. Ultra-detailed textures, realistic skin tones, professional cinematic color grading, sharp focus, HDR, ultra-high resolution 4K, masterpiece quality.",
"style_tags": [
"cinematic",
"night photography",
"urban portrait",
"shallow depth of field",
"bokeh",
"street style",
"HDR",
"professional",
"wet street reflections"
],
"negative_prompt": [
"low quality",
"blurry",
"flat lighting",
"distorted face",
"bad anatomy",
"extra limbs",
"oversharpening",
"text",
"watermark",
"logo",
"daylight"
]
}
Four-Panel Comic Prompt: Gargle (English Version)
youtube-thumbnailThis is a prompt used to generate a four-panel comic strip using Nano Banana Pro, centered around the theme of 'Gargle' (English version).
展开提示词全文
Today's theme is Let's Gargle (English Version)
Grok Instruction for Generating Nano Banana Pro Image Prompts
youtube-thumbnailThis is an instruction given to the Grok LLM to generate a set of five keyframe image prompts suitable for Nano Banana Pro, based on an existing video generation prompt (which is referenced but not included here). This is a meta-prompt for prompt engineering.
展开提示词全文
Please learn the best practices for nano banana pro prompts. I want to create keyframe still images when executing the following video generation prompt. Please think of and display 5 cuts worth of prompts. (Paste video generation prompt here. See reply 3)
Magazine Cover Prompt using Nano Banana Extension
youtube-thumbnailA user demonstrates using a Chrome extension for Nano Banana to easily access various prompts, including a 'magazine cover' prompt, which they used to create a 'Weekly Magazine' style image, suggesting its utility for social media design.
展开提示词全文
I tried creating a "Weekly Magazine" style using the magazine cover prompt
Eyecatch Image Generation from Video Transcript
youtube-thumbnailA user generated an eyecatch image using Nano Banana based on a video transcript, noting that image generation AI might replace manual creation in Canva.
展开提示词全文
Use the transcript of this video as a prompt to have Nano Banana create an eyecatch image
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。