[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tool-Renset--macai":3,"similar-Renset--macai":203},{"id":4,"github_repo":5,"name":6,"description_en":7,"description_zh":8,"ai_summary_zh":8,"readme_en":9,"readme_zh":10,"quickstart_zh":11,"use_case_zh":12,"hero_image_url":13,"owner_login":14,"owner_name":15,"owner_avatar_url":16,"owner_bio":17,"owner_company":18,"owner_location":18,"owner_email":18,"owner_twitter":18,"owner_website":19,"owner_url":20,"languages":21,"stars":29,"forks":30,"last_commit_at":31,"license":32,"difficulty_score":33,"env_os":34,"env_gpu":35,"env_ram":35,"env_deps":36,"category_tags":40,"github_topics":47,"view_count":68,"oss_zip_url":18,"oss_zip_packed_at":18,"status":69,"created_at":70,"updated_at":71,"faqs":72,"releases":102},787,"Renset\u002Fmacai","macai","All-in-one native macOS AI chat application for virtually any AI provider","macai 是一款专为 macOS 系统打造的轻量级原生 AI 聊天客户端，旨在解决用户在 Mac 设备上访问不同人工智能服务时面临的碎片化问题。它将 ChatGPT、Claude、Google Gemini、Ollama 等主流模型整合到一个统一的界面中，让开发者、研究人员及设计师等用户能够在一个地方高效管理所有对话。macai 的核心优势在于极高的兼容性，支持几乎所有 OpenAI 兼容接口，并允许通过 Ollama 在本地运行开源模型以保护隐私。此外，它还内置了 iCloud 同步功能，确保对话记录与设置能在苹果生态设备间无缝衔接。界面设计极简且支持深色模式，官方承诺不收集任何遥测数据。通过 Homebrew 或官网下载即可快速上手，是 Mac 用户管理多模型对话的得力助手。","\u003Cdiv align=\"center\">\n  \u003Cimg width=\"256\" height=\"256\" alt=\"Macai App icon\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_readme_acab60ac1e29.png\" \u002F>\n\u003C\u002Fdiv>\n\u003Ch2 align=\"center\">macai\u003C\u002Fh2>\n\n\u003Ca href=\"#\">\u003Cimg alt=\"GitHub top language\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flanguages\u002Ftop\u002FRenset\u002Fmacai\">\u003C\u002Fa> \u003Ca href=\"#\">\u003Cimg alt=\"GitHub code size in bytes\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flanguages\u002Fcode-size\u002FRenset\u002Fmacai\">\u003C\u002Fa> \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Factions\u002Fworkflows\u002Fswift-xcode.yml\">\u003Cimg alt=\"GitHub Workflow Status\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002FRenset\u002Fmacai\u002Fswift-xcode.yml\">\u003C\u002Fa> \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fblob\u002Fmain\u002FLICENSE.md\">\u003Cimg alt=\"GitHub\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002FRenset\u002Fmacai\">\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Freleases\u002Flatest\">\u003Cimg alt=\"GitHub all releases\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fdownloads\u002FRenset\u002Fmacai\u002Ftotal\">\u003C\u002Fa>\n\nmacai (macOS AI) is a simple yet powerful native macOS AI chat client that supports most AI providers: ChatGPT, Claude, xAI (Grok), Google Gemini, Perplexity, Ollama, OpenRouter, and almost any OpenAI-compatible APIs.\n\n\u003Cimg width=\"1152\" height=\"821\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_readme_eccb662919ea.png\" \u002F>\n\n## Table of Contents\n- [Downloads](#downloads)\n  - [Manual](#manual)\n  - [Homebrew](#homebrew)\n- [Contributions](#contributions)\n- [Why macai](#why-macai)\n- [Run with ChatGPT, Claude, xAI or Google Gemini](#run-with-chatgpt-claude-xai-or-google-gemini)\n- [Run with Ollama](#run-with-ollama)\n- [System requirements](#system-requirements)\n- [Project status](#project-status)\n- [Build from source](#build-from-source)\n- [iCloud Sync Configuration](#icloud-sync-configuration)\n- [License](#license)\n\n## Downloads\n\n### Manual\nDownload [latest universal binary](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Freleases), notarized by Apple.\n\n### Homebrew\nInstall macai cask with homebrew:\n`brew install --cask macai`\n\n\n## Contributions\nContributions are welcome. Take a look at [Issues page](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues) to see already added features\u002Fbugs before creating new one. \nYou can also support project by funding. This support is very important for me and allows to focus more on macai development.\n\n\u003Ca href=\"https:\u002F\u002Fwww.buymeacoffee.com\u002Frenset1\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fcdn.buymeacoffee.com\u002Fbuttons\u002Fv2\u002Fdefault-yellow.png\" alt=\"Buy Me A Coffee\" style=\"height: 60px !important;width: 217px !important;\" >\u003C\u002Fa>\n\n\n## Why macai\n- **macOS-native and lightweight**\n- **User-friendly**: simple setup, minimalist light\u002Fdark UI\n- **Feature-rich**: vision, image generation, search, reasoning, import\u002Fexport and more\n- **iCloud Sync**: keep chats, messages, and settings in sync across devices\n- **Private and secure**: no telemetry or usage tracking by macai (Note: Apple may collect anonymized telemetry when iCloud Sync is enabled)\n\n\n## Run with ChatGPT, Claude, xAI or Google Gemini\nTo run macai with ChatGPT or Claude, you need to have an API token. API token is like password. You need to obtain the API token first to use any commercial LLM API. Most API services offer free credits on registering new account, so you can try most of them for free.\nHere is how to get API token for all supported services:\n- OpenAI: https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F4936850-where-do-i-find-my-secret-api-key\n- Claude: https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fapi\u002Fgetting-started\n- Google Gemini: https:\u002F\u002Fai.google.dev\u002Fgemini-api\u002Fdocs\u002Fapi-key\n- xAI Grok: https:\u002F\u002Fdocs.x.ai\u002Fdocs#models\n- OpenRouter: https:\u002F\u002Fopenrouter.ai\u002Fdocs\u002Fapi-reference\u002Fauthentication#using-an-api-key\n\nIf you are new to LLM and don't want to pay for the tokens, take a look Ollama. It supports dozens of OpenSource LLM models that can run locally on Apple M1\u002FM2\u002FM3\u002FM4 Macs.\n\n## Run with [Ollama](https:\u002F\u002Follama.com)\nOllama is the open-source back-end for various LLM models. \nRun macai with Ollama is easy:\n1. Install Ollama from the [official website](https:\u002F\u002Follama.com)\n2. Follow installation guides\n3. After installation, select model (llama3.1 or llama3.2 are recommended) and pull model using command in terminal: `ollama pull \u003Cmodel>`\n4. In macai settings, open API Service tab, add new API service (Expert mode) and select type Ollama\":\n   \u003Cimg width=\"607\" height=\"757\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_readme_85c79da849ba.png\" \u002F>\n\n5. Select model, and default AI Assistant and save\n6. Test and enjoy!\n\n## System requirements\nmacOS 14.0 and later (both Intel and Apple chips are supported)\n\n## Project status\nProject is in the active development phase.\n\n### Build from source\n\n#### Option 1: With Apple Developer Account (Full Features)\n\nIf you have an Apple Developer account and want to build with iCloud Sync support:\n\n1. Clone the repository: `git clone https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai.git`\n2. Open `macai.xcodeproj` in Xcode\n3. Select your team in Signing & Capabilities\n4. *(Optional)* To enable iCloud Sync in Debug builds, remove `DISABLE_ICLOUD` from Build Settings → Swift Compiler → Active Compilation Conditions\n5. Build and run\n\n> **Note:** By default, Debug builds have iCloud Sync disabled via the `DISABLE_ICLOUD` flag to simplify contributor setup. Release builds have iCloud Sync enabled.\n\n#### Option 2: Without Apple Developer Account (No iCloud Sync)\n\nIf you don't have an Apple Developer account, you can still build and run the app without iCloud Sync:\n\n**Using Xcode:**\n1. Clone the repository: `git clone https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai.git`\n2. Open `macai.xcodeproj` in Xcode\n3. Select the `macai` target → Build Settings tab\n4. Search for `CODE_SIGN_ENTITLEMENTS`\n5. Change the value from `macai\u002Fmacai.entitlements` to `macai\u002Fmacai-no-icloud.entitlements`\n6. In Signing & Capabilities, set \"Signing Certificate\" to \"Sign to Run Locally\"\n7. Build and run\n\n**Using Command Line:**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai.git\ncd macai\nxcodebuild -scheme macai \\\n  -configuration Debug \\\n  CODE_SIGN_IDENTITY=\"-\" \\\n  CODE_SIGN_ENTITLEMENTS=\"macai\u002Fmacai-no-icloud.entitlements\" \\\n  DEVELOPMENT_TEAM=\"\" \\\n  CODE_SIGNING_ALLOWED=NO \\\n  build\n```\n\n> **Note:** The app built without iCloud entitlements will work normally, but the iCloud Sync feature will not be available. All other features (chat, API services, personas, etc.) will work as expected.\n\n### iCloud Sync Configuration\n\n#### For Contributors (Debug Builds)\n\niCloud Sync is **disabled by default** in Debug builds via the `DISABLE_ICLOUD` compiler flag. This simplifies the development setup and avoids CloudKit-related signing issues for contributors without an Apple Developer account.\n\n**To enable iCloud Sync in Debug builds:**\n1. Select the `macai` target → Build Settings tab\n2. Search for `SWIFT_ACTIVE_COMPILATION_CONDITIONS` (or \"Active Compilation Conditions\")\n3. Remove `DISABLE_ICLOUD` from the value (leaving just `DEBUG`)\n4. Ensure you have proper entitlements and signing configured (see below)\n\n#### For Forks \u002F Custom Builds\n\nIf you want iCloud Sync to work in a fork or custom build, you must use your own CloudKit container:\n\n1. Create a CloudKit container in your Apple Developer account\n2. Enable the iCloud capability for the macai target in Xcode, and add your container\n3. Update the `CloudKitContainerIdentifier` value in `macai\u002FInfo.plist` to your container ID\n4. Ensure your app's bundle identifier matches the one you registered for the container\n5. Remove `DISABLE_ICLOUD` from Active Compilation Conditions if present\n\nIf `CloudKitContainerIdentifier` is missing, the app falls back to the default container.\n\n## License\n[Apache-2.0](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fblob\u002Fmain\u002FLICENSE.md)\n","\u003Cdiv align=\"center\">\n  \u003Cimg width=\"256\" height=\"256\" alt=\"Macai App icon\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_readme_acab60ac1e29.png\" \u002F>\n\u003C\u002Fdiv>\n\u003Ch2 align=\"center\">macai\u003C\u002Fh2>\n\n\u003Ca href=\"#\">\u003Cimg alt=\"GitHub top language\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flanguages\u002Ftop\u002FRenset\u002Fmacai\">\u003C\u002Fa> \u003Ca href=\"#\">\u003Cimg alt=\"GitHub code size in bytes\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flanguages\u002Fcode-size\u002FRenset\u002Fmacai\">\u003C\u002Fa> \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Factions\u002Fworkflows\u002Fswift-xcode.yml\">\u003Cimg alt=\"GitHub Workflow Status\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002FRenset\u002Fmacai\u002Fswift-xcode.yml\">\u003C\u002Fa> \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fblob\u002Fmain\u002FLICENSE.md\">\u003Cimg alt=\"GitHub\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002FRenset\u002Fmacai\">\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Freleases\u002Flatest\">\u003Cimg alt=\"GitHub all releases\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fdownloads\u002FRenset\u002Fmacai\u002Ftotal\">\u003C\u002Fa>\n\nmacai (macOS AI) 是一个简单却功能强大的原生 macOS AI 聊天客户端，支持大多数 AI 提供商：ChatGPT, Claude, xAI (Grok), Google Gemini, Perplexity, Ollama, OpenRouter，以及几乎所有兼容 OpenAI 的 API。\n\n\u003Cimg width=\"1152\" height=\"821\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_readme_eccb662919ea.png\" \u002F>\n\n## 目录\n- [下载](#downloads)\n  - [手动](#manual)\n  - [Homebrew](#homebrew)\n- [贡献](#contributions)\n- [为什么选择 macai](#why-macai)\n- [使用 ChatGPT, Claude, xAI 或 Google Gemini 运行](#run-with-chatgpt-claude-xai-or-google-gemini)\n- [使用 Ollama 运行](#run-with-ollama)\n- [系统要求](#system-requirements)\n- [项目状态](#project-status)\n- [从源码构建](#build-from-source)\n- [iCloud 同步配置](#icloud-sync-configuration)\n- [许可协议](#license)\n\n## 下载\n\n### 手动\n下载 [最新的通用二进制文件](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Freleases)，已通过 Apple 公证。\n\n### Homebrew\n使用 homebrew 安装 macai cask：\n`brew install --cask macai`\n\n\n## 贡献\n欢迎贡献。在创建新问题之前，请先查看 [Issues 页面](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues) 以了解已添加的功能\u002F修复的漏洞。\n您也可以通过资助来支持该项目。这种支持对我非常重要，使我能够更专注于 macai 的开发。\n\n\u003Ca href=\"https:\u002F\u002Fwww.buymeacoffee.com\u002Frenset1\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fcdn.buymeacoffee.com\u002Fbuttons\u002Fv2\u002Fdefault-yellow.png\" alt=\"Buy Me A Coffee\" style=\"height: 60px !important;width: 217px !important;\" >\u003C\u002Fa>\n\n\n## 为什么选择 macai\n- **macOS 原生且轻量级**\n- **用户友好**：设置简单，极简主义浅色\u002F深色 UI\n- **功能丰富**：视觉、图像生成、搜索、推理、导入\u002F导出等\n- **iCloud 同步**：跨设备保持聊天记录、消息和设置同步\n- **私密安全**：macai 无遥测或使用跟踪（注意：启用 iCloud 同步时，Apple 可能会收集匿名遥测数据）\n\n\n## 使用 ChatGPT, Claude, xAI 或 Google Gemini 运行\n要使用 ChatGPT 或 Claude 运行 macai，您需要拥有 API 令牌。API 令牌类似于密码。您需要先获取 API 令牌才能使用任何商业大语言模型 (LLM) API。大多数 API 服务在新账户注册时会提供免费额度，因此您可以免费试用大部分服务。\n以下是获取所有支持服务的 API 令牌的方法：\n- OpenAI: https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F4936850-where-do-i-find-my-secret-api-key\n- Claude: https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fapi\u002Fgetting-started\n- Google Gemini: https:\u002F\u002Fai.google.dev\u002Fgemini-api\u002Fdocs\u002Fapi-key\n- xAI Grok: https:\u002F\u002Fdocs.x.ai\u002Fdocs#models\n- OpenRouter: https:\u002F\u002Fopenrouter.ai\u002Fdocs\u002Fapi-reference\u002Fauthentication#using-an-api-key\n\n如果您是 LLM 新手且不想为令牌付费，请查看 Ollama。它支持数十种可以在 Apple M1\u002FM2\u002FM3\u002FM4 Mac 上本地运行的开源大语言模型。\n\n## 使用 [Ollama](https:\u002F\u002Follama.com) 运行\nOllama 是各种 LLM 模型的开源后端。\n使用 Ollama 运行 macai 很简单：\n1. 从 [官方网站](https:\u002F\u002Follama.com) 安装 Ollama\n2. 遵循安装指南\n3. 安装完成后，选择模型（推荐 llama3.1 或 llama3.2），并在终端中使用命令拉取模型：`ollama pull \u003Cmodel>`\n4. 在 macai 设置中，打开 API 服务标签页，添加新的 API 服务（专家模式）并选择类型为 Ollama：\n   \u003Cimg width=\"607\" height=\"757\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_readme_85c79da849ba.png\" \u002F>\n\n5. 选择模型，默认 AI 助手并保存\n6. 测试并享受！\n\n## 系统要求\nmacOS 14.0 及更高版本（支持 Intel 和 Apple 芯片）\n\n## 项目状态\n项目处于积极开发阶段。\n\n### 从源码构建\n\n#### 选项 1：拥有 Apple 开发者账号（完整功能）\n\n如果您拥有 Apple 开发者账号并希望构建支持 iCloud 同步的版本：\n\n1. 克隆仓库：`git clone https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai.git`\n2. 在 Xcode 中打开 `macai.xcodeproj`\n3. 在 Signing & Capabilities 中选择您的团队\n4. *(可选)* 要在 Debug 构建中启用 iCloud 同步，请从 Build Settings → Swift Compiler → Active Compilation Conditions 中移除 `DISABLE_ICLOUD`\n5. 构建并运行\n\n> **注意：** 默认情况下，Debug 构建通过 `DISABLE_ICLOUD` 标志禁用 iCloud 同步，以简化贡献者设置。Release 构建则启用 iCloud 同步。\n\n#### 选项 2：没有 Apple 开发者账号（无 iCloud 同步）\n\n如果您没有 Apple 开发者账号，您仍然可以构建并运行应用，但无法使用 iCloud 同步：\n\n**使用 Xcode：**\n1. 克隆仓库：`git clone https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai.git`\n2. 在 Xcode 中打开 `macai.xcodeproj`\n3. 选择 `macai` 目标 → Build Settings 标签页\n4. 搜索 `CODE_SIGN_ENTITLEMENTS`\n5. 将值从 `macai\u002Fmacai.entitlements` 更改为 `macai\u002Fmacai-no-icloud.entitlements`\n6. 在 Signing & Capabilities 中，将\"Signing Certificate\"设置为\"Sign to Run Locally\"\n7. 构建并运行\n\n**使用命令行：**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai.git\ncd macai\nxcodebuild -scheme macai \\\n  -configuration Debug \\\n  CODE_SIGN_IDENTITY=\"-\" \\\n  CODE_SIGN_ENTITLEMENTS=\"macai\u002Fmacai-no-icloud.entitlements\" \\\n  DEVELOPMENT_TEAM=\"\" \\\n  CODE_SIGNING_ALLOWED=NO \\\n  build\n```\n\n> **注意：** 未包含 iCloud 权限设置的应用程序将正常工作，但 iCloud 同步功能将不可用。所有其他功能（聊天、API 服务、角色设定等）将按预期工作。\n\n### iCloud 同步配置\n\n#### 针对贡献者（调试构建）\n\n在调试构建中，iCloud 同步（iCloud Sync）默认通过 `DISABLE_ICLOUD` 编译器标志被**禁用**。这简化了开发环境设置，并避免了没有 Apple Developer 账户的贡献者遇到的 CloudKit 相关签名问题。\n\n**要在调试构建中启用 iCloud 同步：**\n1. 选择 `macai` 目标 → 构建设置标签页\n2. 搜索 `SWIFT_ACTIVE_COMPILATION_CONDITIONS`（或“活动编译条件”）\n3. 从值中移除 `DISABLE_ICLOUD`（仅保留 `DEBUG`）\n4. 确保已正确配置 Entitlements（权限）和签名（见下文）\n\n#### 针对分叉版本 \u002F 自定义构建\n\n如果你想在分叉版本或自定义构建中使用 iCloud 同步，必须使用你自己的 CloudKit 容器：\n\n1. 在你的 Apple Developer 账户中创建一个 CloudKit 容器\n2. 在 Xcode 中为 `macai` 目标启用 iCloud 功能，并添加你的容器\n3. 将 `macai\u002FInfo.plist` 中的 `CloudKitContainerIdentifier` 值更新为你的容器 ID\n4. 确保应用的 Bundle Identifier（包标识符）与你为容器注册的标识符匹配\n5. 如果存在，从活动编译条件中移除 `DISABLE_ICLOUD`\n\n如果缺少 `CloudKitContainerIdentifier`，应用将回退到默认容器。\n\n## 许可证\n[Apache-2.0](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fblob\u002Fmain\u002FLICENSE.md)","# macai 快速上手指南\n\nmacai 是一款原生 macOS AI 聊天客户端，支持 ChatGPT、Claude、Ollama 等多种主流 AI 服务。界面简洁，支持 iCloud 同步，适用于本地及云端大模型交互。\n\n## 1. 环境准备\n\n*   **操作系统**：macOS 14.0 或更高版本（支持 Intel 芯片及 Apple Silicon M1\u002FM2\u002FM3\u002FM4）。\n*   **前置依赖**：\n    *   使用 Homebrew 安装需确保已配置 Homebrew。\n    *   若从源码构建，需安装 Xcode 及 Git。\n\n## 2. 安装步骤\n\n推荐使用以下方式之一进行安装：\n\n### 方式一：Homebrew 安装（推荐）\n在终端中执行以下命令：\n```bash\nbrew install --cask macai\n```\n\n### 方式二：手动下载\n访问 GitHub Releases 页面下载最新通用二进制文件（Universal Binary），该版本已通过 Apple 公证：\n[https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Freleases](https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Freleases)\n\n> **注意**：如需从源码构建以启用 iCloud 同步功能，请参考项目文档中的 `Build from source` 章节。\n\n## 3. 基本使用\n\nmacai 支持连接商业 API 或本地部署的开源模型。\n\n### 使用商业 API (ChatGPT, Claude, Gemini 等)\n1.  获取对应服务的 **API Token**（类似密码）。\n    *   OpenAI: [帮助文档](https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F4936850-where-do-i-find-my-secret-api-key)\n    *   Claude: [文档](https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fapi\u002Fgetting-started)\n    *   Google Gemini: [文档](https:\u002F\u002Fai.google.dev\u002Fgemini-api\u002Fdocs\u002Fapi-key)\n    *   xAI Grok: [文档](https:\u002F\u002Fdocs.x.ai\u002Fdocs#models)\n    *   OpenRouter: [文档](https:\u002F\u002Fopenrouter.ai\u002Fdocs\u002Fapi-reference\u002Fauthentication#using-an-api-key)\n2.  打开 macai，进入设置添加新的 API Service，输入 Token 即可开始对话。\n\n### 使用本地模型 (Ollama)\n适合不想付费或希望数据完全本地的用户。\n1.  **安装 Ollama**：前往官网下载安装 [ollama.com](https:\u002F\u002Follama.com)。\n2.  **拉取模型**：在终端运行以下命令（推荐 llama3.1 或 llama3.2）：\n    ```bash\n    ollama pull \u003Cmodel>\n    ```\n3.  **配置 macai**：\n    *   打开 macai 设置，进入 **API Service** 标签页。\n    *   选择 **Expert mode**，添加新服务并选择类型为 **Ollama**。\n    *   选择刚才拉取的模型，保存后测试连接即可使用。","一位自由软件开发者日常需要在不同 AI 模型间切换进行代码审查和技术文档撰写，同时希望严格保护本地隐私数据。\n\n### 没有 macai 时\n- 需要频繁切换浏览器标签页访问不同平台（如 ChatGPT、Claude、Gemini），操作割裂且容易分心。\n- 每次都要重新输入 API Key 或登录账号，管理多个订阅账户的密钥非常繁琐且不安全。\n- 聊天记录分散在不同网页中，无法在 iPhone 上继续之前的对话，上下文丢失严重。\n- 网页版占用大量内存，导致 Xcode 等开发工具运行变慢，影响整体工作效率。\n\n### 使用 macai 后\n- macai 在一个原生窗口内集成所有模型，无需切换应用即可快速对比不同模型的代码建议。\n- 一次性配置好 API Token，并支持 Ollama 本地部署，敏感代码片段可直接离线处理保障安全。\n- 通过 iCloud 自动同步聊天历史和设置，手机和 Mac 无缝衔接工作流，随时续写思路。\n- 轻量级客户端不占系统资源，配合极简深色模式让夜间编码体验更加舒适流畅。\n\nmacai 通过统一接口和跨设备同步，极大提升了多模型协作的效率和隐私安全性。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FRenset_macai_acab60ac.png","Renset","Renat","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002FRenset_3adfeb57.jpg","Making handy macOS apps",null,"renset.dev","https:\u002F\u002Fgithub.com\u002FRenset",[22,26],{"name":23,"color":24,"percentage":25},"Swift","#F05138",99.7,{"name":27,"color":18,"percentage":28},"Rich Text Format",0.3,847,68,"2026-04-04T01:46:41","Apache-2.0",1,"macOS","未说明",{"notes":37,"python":38,"dependencies":39},"原生 macOS 应用，无需 Python 环境。仅支持 macOS 14.0 及以上版本，兼容 Intel 和 Apple 芯片。云端对话需配置各服务商 API Key，本地推理需配合 Ollama 使用。若从源码构建，需安装 Xcode，iCloud 同步功能需 Apple Developer 账号配置签名。","无需",[],[41,42,43,44,45,46],"开发框架","其他","图像","插件","Agent","语言模型",[48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67],"ai","bot","chat","chatgpt","chatgpt-api","macos","swift","swiftui","api-client","llm","ollama","claude","gemini","google","grok","xai","swiftui-app","xai-api","deepseek","openrouter",3,"ready","2026-03-27T02:49:30.150509","2026-04-06T05:32:27.564262",[73,78,83,88,92,97],{"id":74,"question_zh":75,"answer_zh":76,"source_url":77},3378,"提示 \"Model not found\" 或连接测试失败怎么办？","确保模型已在 Ollama 中下载。请在 macOS 终端运行命令 `ollama pull \u003C模型名>`（例如 `ollama pull qwen2.5`）。模型下载完成后即可通过 Ollama API 在 macai 中使用。","https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues\u002F32",{"id":79,"question_zh":80,"answer_zh":81,"source_url":82},3379,"无法连接到 Ollama 服务器怎么办？","尝试将设置中的 `localhost` 替换为显式 IPv4 地址 `127.0.0.1`。例如将 API 地址 `http:\u002F\u002Flocalhost:11434\u002Fapi\u002Fchat` 修改为 `http:\u002F\u002F127.0.0.1:11434\u002Fapi\u002Fchat`。","https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues\u002F123",{"id":84,"question_zh":85,"answer_zh":86,"source_url":87},3380,"为什么关闭窗口后窗口大小会恢复默认？","这是 SwiftUI 多窗口应用的默认行为。每个窗口独立管理参数，关闭后新建窗口会使用默认值。目前难以更改此行为，因为 SwiftUI 的多窗口机制限制。","https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues\u002F42",{"id":89,"question_zh":90,"answer_zh":91,"source_url":87},3381,"流式输出时文本显示不全或需要手动滚动吗？","自动滚动功能已实现并将在即将发布的版本中更新。建议保持视图在底部以启用自动滚动，当用户向上滚动时该功能会自动停止以避免干扰阅读。",{"id":93,"question_zh":94,"answer_zh":95,"source_url":96},3382,"关闭 Wi-Fi 时应用为什么会崩溃？","该问题在最新版本中可能已被修复或行为已改变。维护者表示无法复现旧版崩溃，但可能会遇到其他流式传输相关的错误提示。建议更新至最新版本。","https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues\u002F66",{"id":98,"question_zh":99,"answer_zh":100,"source_url":101},3383,"删除聊天记录时应用经常崩溃怎么办？","该问题已在 v2.0.0-beta.1 版本中修复。请更新到该版本或更高版本以解决侧边栏删除聊天时的崩溃问题。","https:\u002F\u002Fgithub.com\u002FRenset\u002Fmacai\u002Fissues\u002F24",[103,108,113,118,123,128,133,138,143,148,153,158,163,168,173,178,183,188,193,198],{"id":104,"version":105,"summary_zh":106,"released_at":107},102951,"v2.4.2","\u003Ch3>Fixes for draft handling, iCloud sync, and app crashes\u003C\u002Fh3>\r\n\u003Cp>\r\nThis version includes a couple of fixes for issues related to draft sync and app crash on chat deletion. Also, iCloud sync was optimized.\r\n\u003C\u002Fp>","2026-01-16T20:50:17",{"id":109,"version":110,"summary_zh":111,"released_at":112},102952,"v2.4.0","\u003Ch3>Introducing support for PDF uploads and image improvements, and more\u003C\u002Fh3>\r\n\u003Cp>\r\nThis version includes a number of improvements and fixes:\r\n\u003Cul>\r\n\u003Cli>[NEW] Support for PDF and image attachments has been added to the OpenRouter API service. All OpenRouter models support PDF attachments. Image attachments are supported for select models\u003C\u002Fli> \r\n\u003Cli>[NEW] Support for PDF attachments has been added for OpenAI and Gemini API services.\u003C\u002Fli>\r\n\u003Cli>[NEW] Shortcut Cmd+Shift+Delete to clear opened chat (thanks @boriselec)\u003C\u002Fli>\r\n\u003Cli>[Improvement] The image previews have been completely overhauled to use the native macOS Quick Preview with the Share, Gallery, and \"Open in Preview\" options\u003C\u002Fli>\r\n\u003Cli>[Improvement] The attached files are now saved as a draft until the message is sent\u003C\u002Fli>\r\n\u003Cli>[Improvement] Reasoning text is now collapsed by default\u003C\u002Fli>\r\n\u003Cli>[Improvement] Reasoning time is now displayed and saved\u003C\u002Fli>\r\n\u003Cli>[Fix] The bug that caused the sidebar with the chat list to automatically expand when a new chat was created has been fixed. It now only expands when the first chat is created\u003C\u002Fli>\r\n\u003Cli>Other minor fixes and improvements\u003C\u002Fli>\r\n\u003C\u002Ful>\r\n\u003Cimg width=\"500\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F4b700293-0c8e-45ad-a1b2-82bc21701114\" \u002F>\r\n\r\n\u003C\u002Fp>","2026-01-14T20:37:21",{"id":114,"version":115,"summary_zh":116,"released_at":117},102953,"v2.3.3","\u003Ch3>Fix Gemini API service\u003C\u002Fh3>\r\n\u003Cp>\r\nThis version resolves a bug that caused the Gemini models to miss the previous chat context.\r\n\u003C\u002Fp>","2026-01-08T22:07:06",{"id":119,"version":120,"summary_zh":121,"released_at":122},102954,"v2.3.2","\u003Ch3>Minor update with new icon and couple of bug fixes\u003C\u002Fh3>\r\n\u003Cp>\r\nThis version introduces an icon update to meet macOS 26 requirements, including a dark version. It also fixes a bug that prevented system messages from being edited in chats and improves internal application update handling. Additionally, it addresses a small UI bug for empty application states.\r\n\u003C\u002Fp>\r\n\u003Cp>\r\nHappy New Year to all macai users!  🎉\r\n\u003C\u002Fp>","2026-01-01T19:42:23",{"id":124,"version":125,"summary_zh":126,"released_at":127},102955,"v2.3.1","\u003Ch3>Stop inference button, fix for frozen input on lengthy text, and other fixes\u003C\u002Fh3>\r\n\u003Cp>\r\nThis version introduces an important fix for text input freezing when pasting or typing lengthy text. The Stop Inference (Response) button has also been added. Other changes:\r\n\u003Cul>\r\n\u003Cli>Push notification when new message is received in the background\u003C\u002Fli> \r\n\u003Cli>Unread chats are marked in the chat list and Dock icon\u003C\u002Fli>\r\n\u003Cli>Minor bug fixes\u003C\u002Fli>\r\n\u003C\u002Ful>\r\n\u003C\u002Fp>","2025-12-26T13:37:44",{"id":129,"version":130,"summary_zh":131,"released_at":132},102956,"v2.3.0","\u003Ch3>Introducing support for iCloud sync, full offline backups, and more\u003C\u002Fh3>\r\n\u003Cp>\r\n\u003Cul>\r\n\u003Cli>This version adds iCloud synchronization support between devices, which is useful if you use macai on more than one Mac.\u003C\u002Fli> \r\n\u003Cli>It's now possible to create and export\u002Fimport full local backups.\u003C\u002Fli>\r\n\u003Cli>Chats now open much faster.\u003C\u002Fli>\r\n\u003Cli>New official website: \u003Ca href=\"https:\u002F\u002Fmacai.chat\">https:\u002F\u002Fmacai.chat\u003C\u002Fa>\r\n\u003C\u002Ful>\r\n\u003C\u002Fp>","2025-12-19T14:57:20",{"id":134,"version":135,"summary_zh":136,"released_at":137},102957,"v2.2.5","\u003Ch3>Introducing support for the Gemini 3 Pro and Nano Banana 3 Pro\u003C\u002Fh3>\r\n\u003Cp>This version adds support for the new models recently released by Google. You can change the model by editing an existing Gemini API service or adding a new one.\u003C\u002Fp> \r\n\u003Cp>Claude Opus 4.5 is also now available in the 'Add Service' pop-up.\u003C\u002Fp>","2025-11-24T20:43:57",{"id":139,"version":140,"summary_zh":141,"released_at":142},102958,"v2.2.4","\u003Ch3>Minor update with bug fixes\u003C\u002Fh3>\r\n\u003Cp>This version fixes a couple of minor yet annoying bugs:\u003C\u002Fp>\r\n\u003Cul>\r\n    \u003Cli>Bug where some changes like model in the API service were not being applied to the chat until the API service was selected again\u003C\u002Fli>\r\n    \u003Cli>Bug where user's message is duplicated when using \"Retry\" operation\u003C\u002Fli>\r\n\u003C\u002Ful>","2025-11-22T23:13:36",{"id":144,"version":145,"summary_zh":146,"released_at":147},102959,"v2.2.3","### Introducing support for the OpenAI Responses API and image generation\r\nmacai now supports the modern [OpenAI Responses API](https:\u002F\u002Fplatform.openai.com\u002Fdocs\u002Fapi-reference\u002Fresponses). This has enabled me to finally add support for OpenAI's image generation. More features and tools based on the OpenAI Responses API will be added in future releases.\r\n\r\n💁‍♂️ Important: if you update macai, all legacy OpenAI Completions API services (i.e. those with the default URL https:\u002F\u002Fapi.openai.com\u002Fv1\u002Fchat\u002Fcompletions) will automatically convert to the new Responses API the first run. If you want to use the old Completions API with your OpenAI account for any reason, you will need to manually change the API service to \"Generic Completions API\".\r\n\r\nAlso, some small bugs were fixed.\r\n\r\n### New contributors\r\nThanks to @ResistanceTo for the small yet important UI polish!","2025-10-21T21:29:17",{"id":149,"version":150,"summary_zh":151,"released_at":152},102960,"v2.2.2","### 🍌 Introducing special release: native Gemini API and Nano Banana image generation support, and more!\r\n\r\nThis release is special because it introduces a few cool new features:\r\n- The native Gemini API is now supported, including vision and image generation.\r\n- Nano Banana by Google Gemini is now fully supported. Generate new images from scratch or edit existing ones — Nano Banana is surprisingly good at this! You can even use a few input images to merge them into one! \r\n\u003Cdetails>\r\n\u003Csummary>Example of image edits\u003C\u002Fsummary>\r\n\u003Ctable>\r\n\u003Ctr>\r\n\u003Ctd>\r\n\u003Cimg height=\"400\" alt=\"CleanShot 2025-09-20 at 20 30 54@2x\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F80c322b8-8671-4b39-9b65-553c611e19af\" \u002F>\r\n\u003C\u002Ftd>\r\n\u003Ctd>\r\n\u003Cimg height=\"400\" alt=\"CleanShot 2025-09-20 at 20 30 03@2x\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fb3eef0c0-9781-49c6-8b2a-1c2251f65573\" \u002F>\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003Ctr>\r\n\u003Ctd>\r\nGenerate new image based on the provided one\r\n\u003C\u002Ftd>\r\n\u003Ctd>\r\nMerge 2 images into one\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003C\u002Ftable>\r\n\u003C\u002Fdetails>\r\n\r\nAnother great use case is the old photo restoration preserving photo details:\r\n\u003Cdetails>\r\n\u003Csummary>Photo restoration and colorization\u003C\u002Fsummary>\r\n\u003Cimg height=\"400\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fe63a5354-8b7d-4a44-898d-3e37b08926db\" \u002F>\r\n\u003C\u002Fdetails>\r\n\r\nTo simplify the setup process, I have also introduced a new, user-friendly method for adding the most popular models. The legacy interface also persists for those who need it.\r\n\r\n\u003Cdetails>\r\n\u003Csummary>New Add Service UI\u003C\u002Fsummary>\r\n\u003Cimg height=\"400\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fe1b7c973-716b-441c-97ec-64c7cde59826\" \u002F>\r\n\u003C\u002Fdetails>\r\n\r\nA couple of minor bugs were also resolved.\r\n- A rare bug that prevented attached images from being sent to the server.\r\n- A bug that caused duplicate API Services\u002FAI Assistant items to appear after adding new ones.\r\n- A bug that occurred when the message content was copied to the message input on Retry.\r\n- Other minor bugs that I discovered during testing.\r\n\r\nRenat","2025-09-20T22:26:10",{"id":154,"version":155,"summary_zh":156,"released_at":157},102961,"v2.2.1","This build includes an important fix for a bug that could cause parts of messages to be silently truncated and **recommended for all users**. This issue mostly affected messages with lists and was caused by the default behaviour of the Text() view in SwiftUI (thanks, SwiftUI!).\r\n\r\nI've also included a couple of minor fixes in the API Service Editor.","2025-09-05T22:33:31",{"id":159,"version":160,"summary_zh":161,"released_at":162},102962,"v2.2.0","### Introducing the new search function (beta), end of macOS 13 support, and added macOS 26 support!\r\n\r\n\u003Cimg width=\"597\" height=\"364\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F28382dca-32af-4e51-8aac-3422f3ff77a4\" \u002F>\r\n\r\nIt's been a long time, and I'm finally happy to release a new version which includes a fully featured search function across all chats. You can search inside an open chat and navigate between occurrences. \r\n\r\nI've also made a bunch of small improvements and fixes, such as faster chat opening, fixing app crashes when deleting chats, and autoscroll fixes. \r\n\r\nMore importantly, I've added support for the new macOS 26 (surprisingly, the macOS 26 build with the new UI had some minor issues, and I have tried to find and fix them all).\r\n\r\nAlso, I've decided to end the macOS 13 support - SwiftUI is a developing framework and has a lot of limitations for older macOS target, and we need to move forward.\r\n\r\nI have big plans for future releases and hope to find enough time to implement them soon. Stay tuned and connected!\r\n\r\nRenat","2025-09-01T20:49:32",{"id":164,"version":165,"summary_zh":166,"released_at":167},102963,"v2.1.1","\u003Ch1>v2.1.1\u003C\u002Fh1>\r\n\u003Cdiv>\r\n    \u003Ch2>Improvement\u003C\u002Fh2>\r\n    \u003Cul>\r\n        \u003Cli>Better Markdown support (for headers and quotes)\u003C\u002Fli>\r\n    \u003C\u002Ful>\r\n\u003C\u002Fdiv>","2025-07-08T20:35:41",{"id":169,"version":170,"summary_zh":171,"released_at":172},102964,"v2.1.0","## Features \r\n* ⚡️ Finally added support for uploading images (OpenAI and Google Gemini services are supported) \r\n* Performance improvements\r\n \r\n## Fixes \r\n* Fixed handling of really, really large messages (thousands lines)\r\n* Fixed bug with disappearing text selection on moving cursor out of message bubble ","2025-04-07T23:17:19",{"id":174,"version":175,"summary_zh":176,"released_at":177},102965,"v2.0.11","## Improvements\r\n* ⚡️ Add ability to pin chats\r\n* ⚡️ Add support for OpenRouter API service with hundreds of models, including many free models such as **deepseek-r1** and **gemini-2.0-flash-thinking-exp!**\r\n* In API service settings, models are sorted alphabetically now","2025-03-04T20:53:46",{"id":179,"version":180,"summary_zh":181,"released_at":182},102966,"v2.0.10","## Improvements\r\n* ⚡️ Add support for Deepseek API service","2025-03-03T22:17:25",{"id":184,"version":185,"summary_zh":186,"released_at":187},102967,"v2.0.9","## Improvements\r\n* ⚡️ Faster chat opening (also removed scrolling animation)\r\n* ⚡️ Improved chat list performance\r\n* Auto-scrolling during message stream\r\n* Better fonts for code blocks ([Fira Code](https:\u002F\u002Fgithub.com\u002Ftonsky\u002FFiraCode) with nice ligatures is now used by default)\r\n\r\n## Fixes\r\n* Fix support for OpenAI's o3 models","2025-02-16T23:22:37",{"id":189,"version":190,"summary_zh":191,"released_at":192},102968,"v2.0.8","## Improvements\r\n* ⚡️ Allow to change font size and theme in Settings:\r\n\u003Cimg src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F0e26a548-da44-4938-ba30-a2b96740e78a\" width=\"624\" \u002F>\r\n\r\n## Fixes\r\n* Don't change AI Assistant in non-empty chats when changing API Service\r\n* Fix custom model control in API service settings","2025-02-08T17:30:47",{"id":194,"version":195,"summary_zh":196,"released_at":197},102969,"v2.0.7","## Improvements\r\n* ⚡️ Add Retry button for last message\r\n* Fetch supported models list for API services: OpenAI, Claude, xAI, Google Gemini\r\n* Update hardcoded models list for Perplexity\r\n* Rename \"Thinking process\" into \"Reasoning\" ","2025-02-07T20:47:54",{"id":199,"version":200,"summary_zh":201,"released_at":202},102970,"v2.0.6","## Improvements\r\n* Force check for new updates on app launch (if enabled). Add setting for automatic update checks\r\n* UI polishing of settings\r\n* Allow changing the AI Assistant ordering via drag-n-drop in settings\r\n* Add button to duplicate API service","2025-02-02T23:32:44",[204,212,221,229,237,248],{"id":205,"name":206,"github_repo":207,"description_zh":208,"stars":209,"difficulty_score":68,"last_commit_at":210,"category_tags":211,"status":69},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[41,43,45],{"id":213,"name":214,"github_repo":215,"description_zh":216,"stars":217,"difficulty_score":218,"last_commit_at":219,"category_tags":220,"status":69},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[41,45,46],{"id":222,"name":223,"github_repo":224,"description_zh":225,"stars":226,"difficulty_score":218,"last_commit_at":227,"category_tags":228,"status":69},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[41,43,45],{"id":230,"name":231,"github_repo":232,"description_zh":233,"stars":234,"difficulty_score":218,"last_commit_at":235,"category_tags":236,"status":69},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[41,46],{"id":238,"name":239,"github_repo":240,"description_zh":241,"stars":242,"difficulty_score":218,"last_commit_at":243,"category_tags":244,"status":69},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[43,245,246,44,45,42,46,41,247],"数据工具","视频","音频",{"id":249,"name":250,"github_repo":251,"description_zh":252,"stars":253,"difficulty_score":68,"last_commit_at":254,"category_tags":255,"status":69},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[45,43,41,46,42]]