博客

  • 【开源推荐】Cherry Studio:47K+ Stars!多模型AI桌面客户端,一个应用搞定所有大模型

    【开源推荐】Cherry Studio:47K+ Stars!多模型AI桌面客户端,一个应用搞定所有大模型

    🍒 项目名称:Cherry Studio

    GitHub Stars:47.1K+

    🏷️ 开源协议:AGPL-3.0

    💻 支持平台:Windows / macOS / Linux

    🔗 GitHubgithub.com/CherryHQ/cherry-studio

    🌐 官网cherry-ai.com

    📝 项目简介

    Cherry Studio 是一款支持多种大语言模型提供商的开源AI桌面客户端,覆盖 Windows、Mac、Linux 三大平台。它提供智能聊天、自主智能体、300+ 预配置助手能力,可统一访问前沿大语言模型,是2026年最值得关注的 AI 生产力工具之一。

    无论你是用 OpenAI Gemini Claude 等云端模型,还是用 Ollama 跑本地模型,Cherry Studio 都能一站式搞定,告别在多个网页和客户端之间来回切换的烦恼。

    Cherry Studio 界面预览

    ⚙️ 安装要求和过程

    环境要求

    • 操作系统:Windows 10+、macOS 11+、主流 Linux 发行版
    • 无需额外环境配置:开箱即用,下载安装包直接运行
    • 网络要求:使用云端模型需配置 API Key;使用本地模型需提前安装 Ollama 或 LM Studio

    快速安装步骤

    1. 访问 GitHub Releases 页面,下载对应系统的安装包
    2. Windows:下载 .exe 安装包,双击运行
    3. macOS:下载 .dmg 文件,拖入 Applications 文件夹
    4. Linux:下载 .AppImage.deb / .rpm 包安装
    5. 启动后,在设置中配置模型 API Key 或连接本地 Ollama

    🌟 核心功能

    🤖 多 LLM 提供商支持

    支持 OpenAI、Gemini、Anthropic、Claude 等主流云端大模型,同时支持 Ollama、LM Studio 等本地模型,一个客户端搞定所有模型。

    🧠 300+ 预配置 AI 助手

    内置丰富领域的专业 AI 助手模板,涵盖编程、写作、翻译、分析等场景,开箱即用,也支持自定义助手创建。

    📄 多格式文档处理

    支持文本、图片、Office 文档、PDF 等多格式文件处理,内置 RAG 知识库能力,让 AI 基于你的文档作答。

    🔌 MCP 协议支持

    支持模型上下文协议(MCP)服务器,可扩展 AI 能力边界,接入专业工具和数据源。

    🎨 可视化与开发工具

    支持 Mermaid 图表可视化、代码语法高亮、全局搜索、话题管理系统,满足专业用户的需求。

    Cherry Studio 功能展示

    🚀 典型使用场景

    场景一:多模型对比测试

    研究者或开发者需要同时对比 GPT-4.1、Claude 4 Opus、Gemini 2.5 Pro 在同一任务上的表现?Cherry Studio 支持多模型同时对话,同一条消息发送给多个模型,结果并排展示,轻松找出最强模型。

    场景二:本地 + 云端混合使用

    日常聊天用免费的本地模型(Ollama + Qwen3),重要任务切换云端前沿模型。Cherry Studio 统一入口管理,无需在多个工具间切换,数据隐私和模型能力兼得。

    场景三:企业知识库问答

    上传公司文档、技术手册到 Cherry Studio 知识库,结合 RAG 技术,让 AI 基于企业内部知识精准作答,打造专属的企业 AI 助手。

    💡 推荐理由

    作为一款开源免费的 AI 桌面客户端,Cherry Studio 最大的价值在于统一——统一了模型入口、统一了对话历史、统一了工具生态。

    在此之前,用 OpenAI 要去网页,用 Claude 要开另一个网页,本地 Ollama 又是命令行,多模型对比更是麻烦。Cherry Studio 把这些全部整合到一个窗口里,还加上了知识库、MCP 工具扩展、300+ 助手模板,真正做到了”一个客户端,所有 AI 能力”。

    尤其值得一提的是它的开箱即用体验——无需配置 Python 环境,无需折腾依赖,下载安装包,点几下鼠标配置 Key,马上就能用。对非技术用户非常友好。

    如果你同时满足以下任意一点,强烈推荐试试 Cherry Studio:

    • 同时使用多个 AI 模型(GPT/Claude/Gemini/本地模型)
    • 希望 AI 能读取并处理你的本地文档
    • 想要一个统一、美观、功能完整的 AI 客户端
    • 重视数据隐私,希望本地运行 AI

    📥 下载地址

    🍒 GitHub 仓库
    🌐 官方网站
    📦 下载安装包


    📌 本文由 WorkBuddy AI 自动化任务 定时发布,选题自 GitHub 热门 AI 开源项目。欢迎关注本站「开源项目」栏目,每周精选优质 AI 开源工具深度介绍。

  • NotebookLM悄悄升级Gemini 3.5,你的AI笔记本现在能跑代码了

    NotebookLM Gemini 3.5 upgrade
    NotebookLM升级Gemini 3.5,每一个笔记本现在都连着一台云电脑 | 图源:The Verge

    你的AI笔记本,现在能自己跑代码了

    谷歌这两天悄悄给 NotebookLM 推了个大版本更新,可能很多人还没注意到。简单说就是:NotebookLM 现在用上了 Gemini 3.5,而且每一个笔记本背后都连着一台”云电脑”。

    这个变化比听起来要大。以前 NotebookLM 主要做的是帮你整理笔记、回答你导入材料里的问题,本质上是个”阅读理解”工具。现在不一样了——它能写代码、跑代码、生成图表、输出各种格式的文件,已经有点像个住在你笔记本里的研究员了。

    谷歌在官方博客里说,升级后的NotebookLM”更准确、更可靠”。但真正有意思的不是这个说法,而是它现在能做的事。


    云电脑是什么意思

    这次更新最核心的变化,是 NotebookLM 现在跑在谷歌的代理编码平台 Antigravity 上。每一个你创建的笔记本,背后都挂着一台安全的云计算机。

    这意味着什么?意味着 NotebookLM 不再只是”回答问题”——它可以主动写代码来做分析,然后把代码扔到那台云电脑上跑,再把结果返回给你。你要它帮你分析一组数据,它可以直接写Python代码、跑统计、画图表,然后把图表给你。

    以前要做到这个程度,你得自己会写代码,或者至少会用 Colab、Jupyter 这类工具。现在你在 NotebookLM 里用自然语言描述你想做什么,它自己就把代码写了、跑了、结果也给你了。

    输出的格式也多了很多。除了以前的文字回复,现在还能直接给你出 PDF、Excel、PowerPoint、CSV 表格,甚至是数据可视化图片(PNG、SVG)和用 Nano Banana 生成的图像。


    不用导入资料也能用了

    另外一个挺实用的更新:你现在不需要先导入笔记或YouTube视频,直接问 NotebookLM 一个问题,它就会自己去调用谷歌搜索帮你找来源。

    这其实是对之前”发现”功能的升级。以前那个功能主要是帮你找资料来源,现在直接整合进了主流程——你问问题,它找资料,然后基于找到的资料回答你,同时把来源列出来,你可以选择把哪些来源导入到笔记本里。

    这个改动看起来小,实际上降低了使用门槛。以前你得先有组织笔记的习惯,才会想到用 NotebookLM。现在哪怕你只是想快速了解一个话题,也可以直接把它当搜索工具来用,而且比普通搜索靠谱——因为它会给出来源,你可以追溯。


    目前只有高价用户能用

    当然,谷歌一贯的作风——好功能先给有钱的用户用。这次更新目前只推送给 Google AI Ultra 订阅用户和 Workspace 用户。普通用户什么时候能用上,谷歌没说。

    AI Ultra 每月要付的钱不少,Workspace 也是企业级收费。所以某种程度上,这也是谷歌在给自己的高价订阅套餐增加卖点。但你不得不承认,如果这些功能最终下放到免费版或低价版,竞争力会相当强。

    想想看:一个能自己跑代码来做分析的 AI 笔记工具,还自带搜索和来源追溯,而且所有计算都在云端完成,不占用你本地资源。对于做研究、写报告、分析数据的人来说,这基本上就是把以前需要好几个工具配合才能做的事,塞进了一个界面里。

    • 研究人员可以用它快速做文献综述和数据分析
    • 学生可以用它整理学习笔记、生成复习材料
    • 职场人可以用它处理报告、做数据可视化
    • 内容创作者可以用它整理素材、生成多格式内容

    AI笔记工具的角逐开始了

    NotebookLM 这个产品方向,其实挺能代表谷歌在 AI 消费级产品上的思路:不追求最炫的演示,而是把 AI 能力悄无声息地嵌进你已经在用的工具里。

    对比一下,苹果的 WWDC 刚说要让 Siri 能帮你做更多事,微软在推 Copilot,Notion 也有 AI 功能。但 NotebookLM 的打法不太一样——它从一个明确的场景(笔记+研究)切入,然后把 AI 能力一层层加上去,每一步都让你觉得”嗯,这个确实有用”,而不是”哇好厉害但我要它干嘛”。

    这次加上云电脑和代码执行能力,算是往前迈了一大步。接下来就看其他家怎么跟了。可以预期的是,类似的”AI+代码执行+多格式输出”的能力,很快就会出现在其他笔记和生产力工具里。

    如果你有 Google AI Ultra 或 Workspace 账号,现在就可以去试试新版本的 NotebookLM。没有的话,可能还得等一段时间。但不管怎样,这个方向值得关注——它很可能改变很多人做研究和整理信息的方式。

  • 微软开源工具被黑,AI开发者的密码正在裸奔

    Microsoft open source tools hacked
    微软托管在GitHub上的数十个开源项目遭黑客入侵,AI开发者凭证面临泄露风险 | 图源:Getty Images

    你电脑上的AI coding工具,可能正在偷你的密码

    上周,如果你用 Claude Code 或 Gemini 的命令行工具做过开发,那你的密码可能已经被偷了。

    事情是这样的。微软托管在 GitHub 上的几十个开源项目被黑了,攻击者往代码里注入了恶意软件。这些项目不是普通的 side project,而是和 Azure 云服务、AI 开发工具直接相关的核心仓库——包括 Claude Code、Gemini CLI、VS Code 相关的工具链。

    据安全公司 Cloudsmith 和 OpenSourceMalware 的分析,当用户在 AI 编码工具里打开被入侵的项目时,恶意软件就会悄悄把你的密码和其他敏感凭证发给攻击者。

    这种攻击有个名字,叫”供应链攻击”。简单说,就是黑客不去打你本人,而是打你信任的工具。你信任微软的开源项目,所以你拉取代码、安装工具,一切看起来很正常——直到你的密码已经到了别人手里。

    而这次的目标特别精准:AI 开发者。这类人群通常能访问云系统、客户数据,甚至是生产环境的密钥。对黑客来说,钓到一条鱼,就能进整个池塘。


    微软这次有点丢脸

    事情曝光后,微软把涉及的数十个仓库全部”禁用”了。你在 GitHub 上访问这些项目,会看到一行提示:”由于违反 GitHub 的服务条款,此仓库的访问已被 GitHub 员工禁用。”

    微软发言人 Ben Hope 的说法是:”我们暂时移除了一部分仓库,作为调查潜在恶意内容的一部分。”话说得很委婉,但翻译过来就是:我们被黑了,正在查。

    更尴尬的是,这已经是短短两周内微软第二次出这种事了。5月中旬,微软的开源项目 Durable Task(一个帮开发者构建应用的工具)就被确认遭到入侵。安全研究人员说,这次最新事件很可能是 Durable Task 项目的”二次入侵”——要么是漏洞没修干净,要么是黑客又找到了新的入口。

    不管是哪种情况,都说明一个问题:微软对开源项目的安全防护,没有大家想象的那么结实。


    这件事为什么值得你关心

    你可能会想:我又不是 AI 开发者,关我什么事?

    其实关系比你想象的大。供应链攻击的本质是信任被滥用。今天黑客盯上的是 AI 开发工具,明天可能就是你我都在用的某个 npm 包、某个 Python 库。一旦恶意代码进入上游,顺着依赖链一路往下传,最终受影响的是每一个普通用户。

    而且,AI 开发工具这两年越来越普及。以前只有专业程序员才会碰命令行工具,现在有了 AI 辅助,很多人第一次学着用 Claude Code 或 Cursor 来写代码。用户基数大了,攻击的价值就高了。

    • 如果你最近用过微软旗下的开源 AI 工具,建议去查一下自己的凭证有没有异常登录记录
    • 重要密码该改的改,启用了双重验证的确认一下
    • 从 GitHub 拉取过微软项目的开发者,留意一下仓库状态,别用到被污染的版本

    大厂也不是铜墙铁壁

    这件事最让人意外的,不是开源项目被黑——这种事其实挺常见——而是微软被黑了。作为地球上安全预算最充足的技术公司之一,微软拥有几乎无限的资源来防御此类攻击。但结果证明,资源多不等于不会出事。

    开源生态的复杂性,可能是最大的软肋。一个项目依赖几十个上游包,每个包又依赖更多包,攻击面之大,超出了任何一家公司的掌控能力。黑客只需要找到最薄弱的那一环,就能顺着链条打穿整条线。

    目前微软还没有公布受影响的具体人数。公司只说”通知了少数可能拉取过受影响仓库内容的客户”——这个”少数”是多少,只有他们自己知道。如果你是用微软 AI 工具做开发的,现在去检查一下,总比事后后悔强。

  • Z世代夏日碎片感九宫格拼贴

    Z世代夏日碎片感九宫格拼贴

    Z世代夏日碎片感九宫格拼贴



    🤖 ChatGPT

    🇺🇸 English Prompt

    Ultra-viral Gen Z summer photo dump collage, arranged in a clean 3×3 grid layout against a vibrant electric-blue studio background. A stylish young woman with short dark hair wears oversized black sunglasses, white wireless headphones, a cozy oversized white zip-up sweater, black biker shorts, light blue crew socks, and chunky pastel-blue sneakers.
    Each frame captures a different playful pose and candid moment: jumping mid-air, kicking a sneaker directly toward the camera with dramatic wide-angle perspective, stretching her arms out joyfully, leaning forward for a selfie-style shot, laughing naturally, sitting with oversized shoes dominating the foreground, and dancing with energetic movement. One frame features her holding headphones toward the camera. Another frame is presented inside a white Polaroid-style photo frame taped to the background.
    Hand-drawn doodle elements connect the photos, including white arrows, curved lines, motion marks, stars, hearts, smiley faces, flowers, sparkles, and handwritten notes such as “Hi~”, “Summer Mood”, “Good Vibes”, and “Stay Happy”. Bright daylight studio lighting, cheerful expression, dynamic body language, playful perspective distortion, trendy Pinterest aesthetic, Instagram photo dump style, youth culture, summer energy, clean composition, highly shareable social media content, vibrant colors, sharp focus, professional photography, ultra-detailed, lifestyle fashion editorial, viral visual storytelling.
    Style Keywords:
    Gen Z aesthetic, Pinterest-core, Instagram photo dump, scrapbook collage, doodle art overlay, playful lifestyle photography, vibrant blue backdrop, wide-angle lens, candid energy, trendy social media campaign, youthful happiness, ultra-viral composition.

    🇨🇳 中文提示词

    爆火的Z世代夏季照片碎片感拼贴,排列在整洁的3×3网格布局中,背景是充满活力的电蓝色影棚背景。一位留着黑色短发的时尚年轻女性,戴着超大黑色太阳镜、白色无线耳机,穿着舒适的超大白色拉链毛衣、黑色骑行短裤、浅蓝色长袜和厚底淡蓝色运动鞋。每一帧都捕捉了不同的俏皮姿势和抓拍瞬间:在半空中跳跃、用戏剧性的广角透视直接向镜头踢出运动鞋、愉快地伸展双臂、向前倾身进行自拍风格的拍摄、自然地大笑、坐着且超大的鞋子在前景中占据主导地位、以及带着充满活力的动作跳舞。其中一帧表现她对着镜头拿着耳机。另一帧呈现在贴在背景上的白色拍立得风格照片框内。手绘涂鸦元素连接着照片,包括白色箭头、曲线、运动标记、星星、心形、笑脸、花朵、闪烁的光点,以及“Hi~”、“Summer Mood”、“Good Vibes”和“Stay Happy”等手写笔记。明亮的日光工作室照明、愉快的表情、动态的肢体语言、俏皮的透视畸变、时尚的Pinterest美学、Instagram照片碎片风格、青年文化、夏日能量、干净的构图、极具分享性的社交媒体内容、鲜艳的色彩、清晰的焦点、专业摄影、超精细、生活方式时尚编辑、病毒式视觉叙事。风格关键词:Z世代美学、Pinterest风格、Instagram照片碎片、剪贴簿拼贴、涂鸦艺术覆盖、俏皮的生活方式摄影、鲜艳的蓝色背景、广角镜头、抓拍能量、时尚社交媒体宣传、青春快乐、极具病毒性的构图。
  • 穿阿根廷球衣的年轻人城市公交街拍

    穿阿根廷球衣的年轻人城市公交街拍

    穿阿根廷球衣的年轻人城市公交街拍



    🤖 ChatGPT

    🇺🇸 English Prompt

    Ultra-realistic candid portrait of a young person with short wavy dark hair wearing an oversized Argentina national football jersey (#19), sitting on a city bus and at a bus stop during golden hour. Soft warm sunlight streams through large windows, creating dramatic light-and-shadow patterns across the face. Natural makeup, flawless pale skin, expressive eyes, relaxed confident expression, editorial street photography aesthetic. Urban European city background visible through the windows with bicycles, storefronts, and classic architecture. Shot on a 50mm lens, shallow depth of field, cinematic composition, authentic lifestyle fashion photography, highly detailed skin texture, natural colors, soft highlights, subtle film grain, Instagram editorial style, photorealistic, 8K, high dynamic range.
    
    Negative Prompt:
    
    blurry, low resolution, overexposed, oversaturated, extra limbs, distorted face, bad anatomy, unrealistic skin, heavy makeup, cartoon, CGI, watermark, text, logo artifacts, duplicate features, poor lighting, noise, cropped head, deformed hands.
    
    Settings (optional):
    •Aspect Ratio: 4:5
    •Lens: 50mm f/1.8
    •Lighting: Golden hour sunlight
    •Style: Cinematic street photography, fashion editorial, photorealistic.

    🇨🇳 中文提示词

    超写实的年轻人抓拍肖像,留着深色短卷发,穿着一件大号的阿根廷国家足球队球衣(19号),在黄金时刻坐在城市公交车上和公交车站。柔和温暖的阳光穿过大窗户,在脸上形成戏剧性的光影图案。自然妆容,完美无瑕的苍白皮肤,富有表现力的眼睛,轻松自信的表情,编辑级街头摄影美学。透过窗户可以看到欧洲城市的背景,有自行车、店面和经典建筑。使用50mm镜头拍摄,浅景深,电影感构图,真实的时尚生活摄影,高度详细的皮肤纹理,自然的色彩,柔和的高光,微妙的电影胶片颗粒,Instagram编辑风格,写实摄影,8K,高动态范围。
    
    负面提示词:
    
    模糊,低分辨率,曝光过度,过度饱和,额外的肢体,扭曲的面部,不良解剖结构,不真实的皮肤,浓妆,卡通,CGI,水印,文字,标志伪影,重复的特征,光线差,噪点,剪裁的头部,变形的手。
    
    设置(可选):
    •宽高比:4:5
    •镜头:50mm f/1.8
    •光照:黄金时刻阳光
    •风格:电影级街拍摄影,时尚编辑,写实摄影。
  • 混乱搞怪丑萌马克笔简笔画插画

    混乱搞怪丑萌马克笔简笔画插画

    混乱搞怪丑萌马克笔简笔画插画



    🤖 ChatGPT

    🇺🇸 English Prompt

    Turn this photo into a chaotic funny doodle illustration, intentionally messy and low-skill, as if drawn quickly with a cheap marker, crayon, or worn-out felt pen on paper.
    
    Create exaggerated facial features with awkward proportions, uneven eyes, oversized head, tiny body, crooked smile, and clumsy anatomy while still keeping the person recognizable. Use rough childish sketch lines, shaky hand-drawn strokes, visible scribbles, overlapping outlines, accidental marks, and random doodles around the scene. Add a simple cartoon-style background with badly drawn buildings, trees, clouds, street elements, and uneven perspective. Coloring should look careless and imperfect, with visible stroke texture, inconsistent fill areas, wax crayon texture, marker bleed, and irregular shading. Include playful imperfections like crossed-out lines, unfinished details, random arrows, tiny notes, stars, swirls, and abstract scribbles. Overall aesthetic should feel humorous, spontaneous, handmade, energetic, goofy, and intentionally unpolished, resembling a child's sketchbook mixed with absurd internet meme art. High texture detail, paper grain visible, asymmetrical composition, awkward framing, expressive doodle chaos, raw sketch energy.

    🇨🇳 中文提示词

    将这张照片变成一幅混乱有趣的涂鸦插画,刻意表现得凌乱且低技术力,就像是用便宜的马克笔、蜡笔或磨损的毛毡笔在纸上快速画出来的一样。创造夸张的面部特征,具有尴尬的比例、不均匀的眼睛、超大的头、微小的身体、歪斜的微笑和笨拙的解剖结构,同时仍保持人物的可识别性。使用粗糙的儿童素描线条、颤抖的手绘笔触、可见的乱涂乱画、重叠的轮廓、偶然的标记和场景周围随机的涂鸦。添加一个简单的卡通风格背景,包含画得很糟糕的建筑、树木、云彩、街道元素和不均匀的透视。上色看起来应该粗心且不完美,具有可见的笔触纹理、不一致的填充区域、蜡笔纹理、马克笔晕染和不规则的阴影。包含俏皮的不完美之处,如划掉的线条、未完成的细节、随机的箭头、微小的笔记、星星、漩涡和抽象的乱画。整体美感应该是幽默、自发、手工制作、充满活力、愚蠢且刻意未经修饰的,类似于儿童速写本与荒诞互联网模因艺术的混合体。高纹理细节,纸张颗粒可见,非对称构图,尴尬的框架,富有表现力的涂鸦混乱,原始的素描能量。
  • 复古动物图标文字组合商标设计

    复古动物图标文字组合商标设计

    复古动物图标文字组合商标设计



    🤖 ChatGPT

    🇺🇸 English Prompt

    Create a single original Retro Animal Symbol Wordmark Logo.
    
    Brand name: [BRAND NAME]
    Subtitle: [CATEGORY / TAGLINE]
    Brand type: [coffee shop / pet café / kids music studio / craft whisky / bakery / lifestyle brand, etc.]
    Animal: [otter / cat / finch / bison / fox / dog / rabbit / bird, etc.]
    Brand mood: [warm / playful / elegant / vintage / friendly / premium / cozy / bold]
    Color direction: [muted retro colors, 2–4 colors only]
    Background: [warm cream / soft ivory / muted sage / deep forest green / dark charcoal]
    
    Style requirements:
    - flat vector logo
    - simplified animal symbol
    - retro but clean
    - small-brand identity feeling
    - custom typography, not generic font
    - clear readable brand name
    - balanced logo composition
    - comfortable medium size on the canvas
    - enough negative space
    - 2–4 muted colors only
    - optional small details: stars, dots, short lines, badge elements, year marks
    
    The animal should feel like a brand symbol, not a random illustration.  
    It can have a small action or personality, such as holding a cup, playing music, sitting calmly, walking, stretching, or carrying a tiny object.
    
    Avoid:
    3D, metallic texture, grunge texture, paper grain, realistic shading, overly cute cartoon style, complex background, messy colors, poster layout, sticker look, generic fonts.

    🇨🇳 中文提示词

    创建一个单一的原创复古动物图标文字组合商标。
    
    品牌名称:[BRAND NAME]
    副标题:[CATEGORY / TAGLINE]
    品牌类型:[咖啡馆 / 宠物咖啡馆 / 儿童音乐工作室 / 手工艺威士忌 / 面包店 / 生活方式品牌等]
    动物:[水獭 / 猫 / 雀鸟 / 野牛 / 狐狸 / 狗 / 兔子 / 鸟等]
    品牌氛围:[温暖 / 俏皮 / 优雅 / 复古 / 友好 / 高端 / 舒适 / 大胆]
    色彩方向:[柔和复古色调,仅限 2–4 种颜色]
    背景:[暖奶油色 / 柔和象牙白 / 柔和鼠尾草绿 / 深森林绿 / 深炭灰色]
    
    风格要求:
    - 扁平矢量标志
    - 简化的动物符号
    - 复古且简洁
    - 小众品牌身份感
    - 定制排版,非通用字体
    - 清晰易读的品牌名称
    - 平衡的标志构图
    - 画布上舒适的中等尺寸
    - 足够的负空间
    - 仅限 2–4 种柔和颜色
    - 可选的小细节:星星、点、短线、徽章元素、年份标记
    
    动物应该感觉像一个品牌符号,而不是随机的插图。
    它可以有一个小动作或个性,例如拿一个杯子、演奏音乐、静静地坐着、行走、拉伸或携带一个小物体。
    
    避免:
    3D、金属质感、脏粉质感、纸张纹理、写实阴影、过度可爱的卡通风格、复杂背景、凌乱色彩、海报布局、贴纸外观、通用字体。
  • 黑道千金高级和风写真

    黑道千金高级和风写真

    黑道千金高级和风写真



    🤖 ChatGPT

    🇺🇸 English Prompt

    Please generate a set of 'Yakuza Heiress · High-end Japanese-style Photoshoot' based on the clear single-person portrait reference photo I uploaded, a total of 9 images, all as individual pictures, no nine-square grids, no composite posters. Each 3:4 vertical composition.
    
    ## Fixed General Control
    
    Strictly preserve the real identity characteristics of the reference person: face shape, facial proportions, eye shape, nose, lips, jawline, hairline, skin tone, sense of age, and overall recognizability. No face swapping, do not turn into a stranger, no internet celebrity face, no infantilization, no animation, no AI doll face. Skin retains real texture, slight pores, natural facial shadows, and real photographic quality.
    
    Overall style: Young heir of a Yakuza family, high-end Japanese-style photoshoot, cold and elegant, restrained, dangerous, noble, sense of control, feminine charm but not vulgar. The frame looks like a 10,000-yuan-level studio fashion blockbuster, not cosplay, not a game poster, not cheap studio style.
    
    The scene is unified as a Japanese-style washitsu built in a real photography studio: bright shoji doors, tatami, bamboo shadows, black paper umbrellas, white flower branches, low tea tables, white porcelain tea sets, old folding screens, dark wooden spaces. The frame can be dark, but not pitch black; black fabrics must have layers of gray-black, charcoal black, tea-brown-black, and ink black, with visible folds, embroidery, and sleeve hems.
    
    Lighting: Side-angled hard window light + soft reflected fill light + slight rim light. Shoji doors can be bright, but the person's face, skin, kimono, tea set, and flower branches must not be blurred.
    
    Clothing: High-end black kimono or modified furisode, obvious white inner lining collar, embroidery in dark red, withered gold, old silver, ink green, floral patterns, butterfly patterns, or family crests. Silver-gray-gold brocade obi (waist belt), heavy and three-dimensional. Dark brown-black or black hair, loose or high updo with natural stray hairs, paired with dark red spider lilies, red hairpins, metal hairpins, or fine tassels. Cold and elegant makeup, slender eye makeup, deep red, wine red, or brownish-red lips, no pink girlish vibe.
    
    Body and posture must be real and natural: Only one person appears, normal body proportions, clear logic for arms and legs, joints are natural, fingers are normal. No extra hands or feet, no missing legs, do not show only one leg causing a sense of disconnection, no fused legs, do not twist waist and hips, no toe deformities. In any scene showing legs, the posture of both legs must be understandable: one leg folded, the other extended, or both legs naturally covered by the kimono but the body structure is complete. Actions are high-class and restrained, not pornographic, not excessively sexualized.
    
    Frame restrictions: No text, no vertical Japanese characters, no logos, no watermarks, no photography studio signatures. No firearms, no blood, no fighting. Katana, scabbards, and folding fans can appear, but only as static aura props.
    
    ---
    
    ## 01|Sitting Pose in Tea Room with Paper Umbrella
    
    Inside a Japanese-style washitsu, bright shoji doors are at the rear side; a black paper umbrella, white flower branches, a low tea table, and white porcelain tea sets form the background layers.
    
    The character sits on the tatami, body slightly turned, not sitting upright. One hand holds a tea cup, and the other hand is placed lightly on the knee. The kimono collar is naturally loose, revealing the collarbone and neck lines. She looks directly at the camera with a calm, aloof, and oppressive gaze.
    
    Bust to 3/4 body vertical composition, focusing on the face, neck, tea cup, hands, obi, and the texture of the black kimono.
    
    --- 
    
    ## 02|High-Angle Shot with Folding Fan
    
    A slightly high-angle shot of the complete Japanese-style washitsu, showing tatami texture, window grille light and shadow, low tea table, white porcelain tea set, black paper umbrella, white flower branches, and bamboo shadow screen.
    
    The character sits slightly to one side of the center of the tatami, holding an open black folding fan in one hand, with the other hand naturally supporting on the ground. The posture is elegant and quiet, head down or eyes closed, not looking at the camera, like a moment caught in a snapshot.
    
    The body structure must be stable: both legs must have clear logic. One leg is naturally folded, the other leg extends forward or diagonally forward; the kimono hem naturally spreads out, but it cannot show only one leg causing a sense of disconnection. The focus of the frame is the space, the folding fan, the tatami, the kimono layout, and the body curves.
    
    --- 
    
    ## 03|Close-up with Katana and Gaze
    
    A close-up of the washitsu under bright window grille light, with shoji doors, black paper umbrella outlines, white flower branches, and dark screens in the background.
    
    The character wears a black embroidered kimono, with a clear white collar and an obvious silver-gray-gold obi. Both hands hold a scabbard or a half-drawn katana as a static power prop, without making any attacking moves.
    
    Close-up composition, concentrated on the face, eyes, neck, hands, and sword hilt. She raises her eyes to look directly at the camera, her gaze sharp, cold and elegant, alert, and oppressive. Fingers must be natural, and the sword hilt must not be deformed.
    
    --- 
    
    ## 04|Side Sitting with Extended Long Legs
    
    On the floor of a Japanese-style washitsu, the background has bright shoji doors, bamboo shadows, black paper umbrellas, white flower branches, low tea tables, and white porcelain tea sets. Window grille light hits the tatami and the character's legs at an angle, forming real light and shadow cuts.
    
    The character sits sideways on the tatami, body leaning back slightly, one hand supporting on the ground, the other hand lightly holding a folding fan or near the tea table. Both legs must be complete and natural: one leg folded on one side of the body, the other leg extended forward, showing normal thigh, knee, calf, and foot lines. Do not show only one leg, do not fuse legs, and no foot deformities.
    
    She looks away with her side eyes, her expression lazy, cold, and dangerous. Medium shot vertical composition, highlighting the sense of body extension, waist and hip lines, leg posture, foreground flower branches, and black kimono folds.
    
    --- 
    
    ## 05|Leaning Over the Tea Table
    
    The character is near a low tea table with a white porcelain teapot, tea cups, small flower branches, and a closed folding fan on the table. The background includes shoji doors, black paper umbrellas, and bamboo shadow screens.
    
    She leans forward slightly, one hand holding the edge of the table, and the other hand lightly taking a tea cup or a folding fan. The shoulder line tilts naturally, the waist twists slightly, forming a soft upper body curve. The kimono collar opens naturally, highlighting the collarbone and neck, but maintaining a high-end restraint.
    
    She looks down at the tea cup, then slightly raises her eyes to look at the camera, with an aloof, calm expression carrying a hint of dangerous attraction. Half-body vertical composition, focusing on the face, collarbone, arms, tea set, and tabletop details.
    
    --- 
    
    ## 06|Standing Pose Side View with Katana
    
    In the center of a Japanese-style washitsu, the background has bright shoji doors, bamboo shadow screens, black paper umbrellas, white flower branches, and wooden spaces. Overall more formal, the light is clear but with contrast.
    
    The character stands sideways, weight on one leg, waistline slightly pulled in, and shoulders and neck stretched out. Both hands hold the scabbard or katana vertically; the sword serves as an aura prop, not for attacking. The head turns slightly, and the gaze looks toward the front side or with eyes half-closed, cold and elegant, restrained, and dangerous.
    
    3/4 body vertical composition, focusing on the sword, profile of the face, neck, obi, sleeve hems, and the standing aura. Body proportions must be normal, fingers and sword hilt must not be deformed.
    
    --- 
    
    ## 07|Semi-Lying on Tatami
    
    Dark washitsu floor, background with shoji doors, old screens, low wooden table, white vases, and white flower branches. There are clear window grille light and shadows on the floor, and the foreground can have slightly blurred flower branches or tea sets.
    
    The character is semi-lying on the tatami, one arm supporting the upper body, the other hand lightly placed near the obi or hem. The upper body is slightly twisted, the waistline naturally tightened, and the legs naturally intersect and extend under the kimono.
    
    Note that the body must be real: the positions of both legs must be logical; even if covered by the kimono, it cannot form a missing leg, broken leg, or extra leg. She turns her head to look at the camera, her gaze nonchalant, cold, and charming. Medium-long shot vertical composition, highlighting body curves, tatami texture, light and shadow cuts, and the layers of black fabric.
    
    --- 
    
    ## 08|Looking Back While Half-Kneeling
    
    Dark washitsu, background with bamboo shadow screens, white flower branches, shoji doors, and black paper umbrellas. Oblique light from outside the window illuminates the back, shoulder line, and side face.
    
    The character has her back or side-back to the camera, half-kneeling on the tatami. The patterns, butterfly patterns, and gold embroidery on the back and sleeves of the black kimono are clear, and there is a three-dimensional silver-gold obi knot at the back of the waist.
    
    The body is forward, the head and upper body turn back slightly, forming an elegant arc from the back to the waist. One hand holds the scabbard or folding fan, and the other hand lightly supports the ground or hem. Both legs are in a natural half-kneeling position, covered by the kimono but with complete structure; no strange lower body or extra limbs should appear.
    
    She looks back at the camera, her gaze cold, steady, alert, dangerous yet feminine. 3/4 body vertical composition, focusing on the back, obi, neck, looking-back gaze, and the beauty of the body twist.
    
    --- 
    
    ## 09|Final Converging Main Visual
    
    The main hall of a Japanese family mansion, with shoji doors, dark wooden pillars, old screens, black paper umbrellas, white flower branches, and low tea tables in the background. Overall more solemn and formal, like the cover of the entire photoshoot set.
    
    The character wears the most gorgeous high-end black kimono of the set, with a white inner collar, dark red, withered gold, and old silver embroidery, a heavy silver-gray-gold brocade obi, more obvious spider lily hair accessories, and deep red lips.
    
    She adopts a formal sitting posture with a sense of a queen: sitting upright but with the upper body leaning forward or turning slightly. One hand is placed on the knee, and the other hand holds a tea cup, folding fan, or scabbard. The posture is steady but retains the stretching of the neck, waistline, and feminine aura.
    
    She looks directly at the camera, her gaze calm and with a sense of power, like a Yakuza heiress truly settled into her position. Formal vertical composition, focusing on the face, gaze, neck, hands, obi, and the overall power aura.

    🇨🇳 中文提示词

    请基于我上传的清晰单人人像参考图,生成一组「黑道千金 · 高级和风写真」,共 9 张,全部为单独图片,不要九宫格拼图,不要合成海报。每张 3:4 竖构图。
    
    ## 固定总控
    
    严格保留参考人物本人的真实身份特征:脸型、五官比例、眼睛形状、鼻子、嘴唇、下颌线、发际线、肤色、年龄感和整体辨识度。不能换脸,不能变成陌生人,不能网红脸,不能幼态化,不能动漫化,不能 AI 娃娃脸。皮肤保留真实纹理、轻微毛孔、自然面部阴影和真实摄影质感。
    
    整体风格:黑道家族年轻继承人,高级和风写真,冷艳、克制、危险、贵气、有掌控感,有女性魅力但不低俗。画面像万元级棚拍时尚大片,不是 cosplay,不是游戏海报,不是廉价影楼风。
    
    场景统一为真实摄影棚搭建的日式和室:明亮障子门、榻榻米、竹影、黑色纸伞、白色花枝、低矮茶桌、白瓷茶具、旧屏风、深色木质空间。画面可以暗,但不能死黑;黑色布料必须有灰黑、炭黑、茶褐黑、墨黑层次,能看到褶皱、刺绣和袖摆。
    
    光线:侧向窗格硬光 + 柔和反射补光 + 轻微轮廓光。障子门可以明亮,但人物脸、皮肤、和服、茶具、花枝不能糊。
    
    服装:黑色高级和服或改良振袖,白色内衬领口明显,暗红、枯金、旧银、墨绿、花枝纹、蝶纹或家纹刺绣。银灰金织锦 obi 腰封,厚重立体。深棕黑发或黑发,松散盘发或高盘发,有自然碎发,配暗红彼岸花、红色簪花、金属发簪或细流苏。妆容冷艳,细长眼妆,深红、酒红或棕红唇,不要粉嫩少女感。
    
    身体与姿态必须真实自然:只出现一个人,身体比例正常,双臂双腿逻辑清晰,关节自然,手指正常。不要多手多脚,不要缺腿,不要只露出一条腿造成断裂感,不要融合腿,不要扭曲腰胯,不要脚趾畸形。凡是露腿画面,必须让两条腿的姿态都能被理解:一条腿折起,另一条腿延展,或两条腿都被和服自然遮挡但身体结构完整。动作高级克制,不色情,不过度性感化。
    
    画面限制:不要文字、不要日文竖排字、不要 logo、不要水印、不要摄影机构署名。不要枪支、不要血腥、不要打斗。可以出现武士刀、刀鞘、折扇,但只能作为静态气场道具。
    
    ---
    
    ## 01|纸伞茶室坐姿图
    
    日式和室内,明亮障子门在侧后方,黑色纸伞、白色花枝、低矮茶桌和白瓷茶具形成背景层次。
    
    人物坐在榻榻米上,身体轻微侧转,不要正襟危坐。一只手拿茶盏,另一只手轻放膝上。和服领口自然松弛,露出锁骨和肩颈线条。她直视镜头,眼神冷静、疏离、有压迫感。
    
    半身到 3/4 身竖构图,重点表现脸、肩颈、茶盏、手部、腰封和黑色和服质感。
    
    ---
    
    ## 02|高位俯拍折扇图
    
    略高机位俯拍完整日式和室,能看到榻榻米纹理、窗格光影、低矮茶桌、白瓷茶具、黑纸伞、白色花枝和竹影屏风。
    
    人物坐在榻榻米中央偏一侧,手持打开的黑色折扇,另一只手自然撑地。姿势优雅安静,低头或闭眼,不看镜头,像被抓拍到的一瞬间。
    
    身体结构必须稳定:两条腿都要有清晰逻辑。一条腿自然折起,另一条腿向前或斜前方延展;和服下摆自然铺开,但不能只露出一条腿造成断裂感。画面重点是空间、折扇、榻榻米、和服铺陈和身体曲线。
    
    ---
    
    ## 03|持刀近景眼神图
    
    明亮窗格光下的和室近景,背景有障子门、黑纸伞轮廓、白色花枝和暗色屏风。
    
    人物穿黑色刺绣和服,白色领口清楚,银灰金 obi 明显。双手持一把刀鞘或半出鞘武士刀,作为静态权力道具,不做攻击动作。
    
    近景构图,集中在脸、眼睛、肩颈、手部和刀柄。她抬眼直视镜头,眼神锋利、冷艳、警觉、有压迫感。手指必须自然,刀柄不能变形。
    
    ---
    
    ## 04|侧坐长腿延展图
    
    日式和室地面,背景有明亮障子门、竹影、黑色纸伞、白色花枝、低矮茶桌和白瓷茶具。窗格光斜射在榻榻米和人物腿部,形成真实光影切割。
    
    人物侧坐在榻榻米上,身体微微后仰,一只手撑地,另一只手轻拿折扇或靠近茶桌。两条腿必须完整自然:一条腿折起收在身体一侧,另一条腿向前延展,能看出正常大腿、膝盖、小腿和脚部线条。不要只露出一条腿,不要融合腿,不要脚部畸形。
    
    她侧眼看向远处,神情慵懒、冷淡、危险。中景竖构图,突出身体延展感、腰胯线条、腿部姿态、前景花枝和黑色和服褶皱。
    
    ---
    
    ## 05|茶桌边俯身图
    
    人物靠近低矮茶桌,桌上有白瓷茶壶、茶杯、小花枝和合上的折扇。背景为障子门、黑纸伞和竹影屏风。
    
    她身体微微前倾,一只手扶桌沿,另一只手轻拿茶杯或折扇。肩线自然倾斜,腰部轻微扭转,形成柔和上半身曲线。和服领口自然打开,突出锁骨和肩颈,但保持高级克制。
    
    她低头看向茶杯,再微微抬眼看镜头,表情疏离、冷静、带一点危险吸引力。半身竖构图,重点表现脸、锁骨、手臂、茶具和桌面细节。
    
    ---
    
    ## 06|站姿持刀侧身图
    
    日式和室中央,背景有明亮障子门、竹影屏风、黑色纸伞、白色花枝和木质空间。整体更正式,光线清透但有反差。
    
    人物站立侧身,重心放在一侧腿上,腰线微收,肩颈舒展。双手竖持刀鞘或武士刀,刀作为气场道具,不攻击。头微微偏转,眼神看向侧前方或半垂眼,冷艳、克制、危险。
    
    3/4 身竖构图,重点突出刀、脸部侧影、肩颈、腰封、袖摆和站姿气场。身体比例必须正常,手指和刀柄不能变形。
    
    ---
    
    ## 07|榻榻米半躺图
    
    暗色和室地面,背景有障子门、旧屏风、低矮木桌、白色花瓶和白花枝。地面有清晰窗格光影,前景可有轻微虚化花枝或茶具。
    
    人物半躺在榻榻米上,一侧手臂撑起上半身,另一只手轻搭在腰封或衣摆旁。上身轻微扭转,腰线自然收紧,腿部在和服下自然交错延展。
    
    注意身体必须真实:两条腿的位置要合理,即使被和服遮挡,也不能形成缺腿、断腿或额外腿。她转头看向镜头,眼神漫不经心,冷淡又迷人。中远景竖构图,突出身体曲线、榻榻米纹理、光影切割和黑色布料层次。
    
    ---
    
    ## 08|回头半跪背影图
    
    暗色和室,背景有竹影屏风、白色枝花、障子门和黑色纸伞。窗外斜射光照亮背部、肩线和侧脸。
    
    人物背对或侧背对镜头,半跪坐在榻榻米上。黑色和服背部与袖子上的花纹、蝶纹、金色刺绣清晰,腰后有立体银金色 obi 结。
    
    身体向前,头和上半身微微回转,形成背部到腰部的优雅弧线。一只手握刀鞘或折扇,另一只手轻扶地面或衣摆。双腿为自然半跪坐姿,被和服覆盖但结构完整,不能出现奇怪下半身或多余肢体。
    
    她回头看向镜头,眼神冷、稳、警觉,危险又有女性魅力。3/4 身竖构图,重点表现背部、腰封、肩颈、回头眼神和身体扭转美感。
    
    ---
    
    ## 09|最终收束主视觉
    
    日式家族宅邸主厅,背景有障子门、深色木柱、旧屏风、黑色纸伞、白色枝花和低矮茶桌。整体更庄重正式,像整组写真封面。
    
    人物穿整组最华丽的黑色高级和服,白色内衬领口,暗红、枯金、旧银刺绣,厚重银灰金织锦 obi,彼岸花发饰更明显,深红唇。
    
    她采用女王感正式坐姿:身体端坐但上半身轻微前倾或侧转。一只手放膝上,另一只手扶茶盏、折扇或刀鞘。姿态稳,但保留肩颈舒展、腰线和女性气场。
    
    她直视镜头,眼神冷静、有权力感,像真正坐稳位置的黑道千金。正式竖构图,重点表现脸、眼神、肩颈、手部、腰封和整体权力气场。
  • 高级时尚杂志风格人像大片

    高级时尚杂志风格人像大片

    高级时尚杂志风格人像大片



    🤖 ChatGPT

    🇺🇸 English Prompt

    Please based on this selfie I uploaded, generate a high-end fashion magazine style portrait masterpiece. Retain the person's real facial features, face shape, hairstyle, and personal temperament, but reshape professional studio lighting, a clean high-end background, shallow depth of field, soft skin texture, and cinematic tones. The overall look should be like a magazine cover portrait shot by a professional photographer, with exquisite light and shadow, high-end composition, and the person appearing natural and confident, without looking like excessive skin smoothing or retouching.

    🇨🇳 中文提示词

    请基于我上传的这张自拍照,生成一张高级时尚杂志风格的人像大片。保留人物真实五官特征、脸型、发型和个人气质,但重新塑造专业影棚光线、干净高级背景、浅景深、柔和皮肤质感和电影级色调。整体像专业摄影师拍摄的杂志封面形象照,光影精致,构图高级,人物自然自信,不要像过度磨皮修图。
  • 舒适复古面包店外观动漫少女插画

    舒适复古面包店外观动漫少女插画

    舒适复古面包店外观动漫少女插画



    🤖 ChatGPT

    🇺🇸 English Prompt

    Cozy vintage bakery exterior illustration, anime style, young girl with blue eyes and brown hair in bun wearing blue polka dot dress with white frilly apron and bonnet, holding large wicker basket full of fresh bread and baguettes. Light blue wooden door, cream colored bakery facade, flower boxes overflowing with colorful flowers

    🇨🇳 中文提示词

    舒适的复古面包店外观插画,动漫风格,蓝眼睛、棕色头发扎成发髻的少女,穿着蓝色圆点连衣裙,配有白色荷叶边围裙和软帽,拿着装满新鲜面包和法棍的大柳条篮。浅蓝色木门,奶油色面包店外墙,盛开着五颜六色鲜花的花箱。