<div align="center">

ComfyUI Civitai Gallery<br>_{^{-=SFW-Friendly=-}}

</div>

Changelog (2025-09-11)

• Edit Prompt: A new “Edit Prompt” checkbox has been added to the Civitai Images Gallery. When enabled, it allows users to edit the prompt associated with each image, making it easier to quickly refine or remix prompts in real time. This feature also supports completing and saving prompts for images with missing or incomplete metadata. Additionally, image loading in the Favorites library has been optimized for better performance.

Changelog (2025-09-07)

• 🎬 Video Preview Support: The Civitai Images Gallery now supports video browsing. You can toggle the “Show Video” checkbox to control whether video cards are displayed. To prevent potential crashes caused by autoplay in the ComfyUI interface, look for a play icon (▶️) in the top-right corner of each gallery card. If the icon is present, you can hover to preview the video or double-click the card (or click the play icon) to watch it in its original resolution.

Changelog (2025-09-06)

• One-Click Workflow Loading: Image cards in the gallery that contain ComfyUI workflow metadata will now persistently display a "Load Workflow" icon (🎁). Clicking this icon instantly loads the entire workflow into your current workspace, just like dropping a workflow file. Enhanced the stability of data parsing to compatibly handle and auto-fix malformed JSON data (e.g., containing undefined or NaN values) from various sources, improving the success rate of loading.

• Linkage Between Model and Image Galleries: In the "Civitai Models Gallery" node's model version selection window, a "🖼️ View Images" button has been added for each model version. Clicking this button will now cause the "Civitai Images Gallery" to load and display images exclusively from that specific model version. When in linked mode, the Image Gallery will show a clear notification bar indicating the current model and version being viewed, with an option to "Clear Filter" and return to normal browsing.

Changelog (2025-09-05)

• New Node: Civitai Models Gallery: Added a completely new Civitai Models Gallery node. It allows you to browse, filter, and download models (Checkpoints, LoRAs, VAEs, etc.) directly from Civitai within ComfyUI.

• Model & Resource Downloader: Implemented a downloader for all resource types. Simply click the "Download" button in the new "Resources Used" viewer or the Models Gallery to save files to the correct folders. This requires a one-time setup of your Civitai API key.

• Advanced Favorites & Tagging: The favorites system has been overhauled. You can now add custom tags to your favorite images for better organization.

• Enhanced UI & Workflow Memory: The node now saves all your UI settings (filters, selections, sorting) within your workflow, restoring them automatically on reload.

Overview

ComfyUI Civitai Gallery is a powerful custom node for ComfyUI that integrates a seamless image browser for the Civitai website directly into your workflow. This node allows you to browse, search, and select images from Civitai and instantly import their prompts, negative prompts, and full-resolution images into your workflow. It is designed to significantly speed up your creative process by eliminating the need to switch between your browser and ComfyUI.

The gallery features a fluid, responsive waterfall (masonry) layout that intelligently fills the available space, ensuring a beautiful and efficient browsing experience.

Features

• Direct Civitai Browsing: Browse images from Civitai without leaving the ComfyUI interface.
• Advanced Filtering: Filter images by NSFW level, sort order (Most Reactions, Newest, etc.), time period (Day, Week, etc.), specific tags, and username.
• One-Click Import: Simply click on an image to select it. When you run the workflow, the node will output:
  • positive_prompt: The positive prompt used to generate the image.
  • negative_prompt: The negative prompt.
  • image: The original, full-resolution image.
  • info: A detailed JSON string containing all other available metadata (sampler, steps, seed, model hash, etc.).
• Intelligent Image Loading: The original image is only downloaded if its image output is connected to another node, saving bandwidth and time.
• Stable Waterfall Layout: A fluid, responsive masonry layout that keeps existing images in place when new ones are loaded via infinite scroll.
• Custom UI: Features a custom-styled, permanently visible scrollbar for easy navigation.
• Advanced Lightbox Viewer:
  • Double-click any image to open a full-screen, centered preview.
  • Image Viewer: Supports zooming (mouse wheel) and panning (drag).
  • Gallery Navigation: Use on-screen arrows or keyboard arrow keys (Left/Right) to cycle through all image.

How to Use

1. Add the Node: Press Tab or double-click in your ComfyUI workspace, search for Civitai Gallery, and add the node to your graph.
2. Browse and Filter:
   • Use the dropdown menus and text fields at the top of the node to filter the images according to your needs.
   • Check the "International" box to use, If you are an international network user.
   • Click the "Refresh" button to apply new filters.
   • Scroll down within the gallery to automatically load more images (infinite scroll).
3. Select an Image: Click on any image card in the gallery. A colored border will appear around your selection.
4. Connect the Outputs:
   • Connect the positive_prompt and negative_prompt outputs to the corresponding inputs on your KSampler node or a text display node.
   • Connect the image output to a Preview Image or Save Image node if you need the original image. (Note: The image will only be downloaded if this output is connected).
   • Connect the info output to a Show Text node to view all other generation parameters.
5. Queue Prompt: Run your workflow. The selected image's data will be fed into the connected nodes.

Installation

1. Navigate to your ComfyUI installation directory.
2. Go to the custom_nodes folder.
3. Clone or download this repository into the custom_nodes folder. The final folder structure should be ComfyUI/custom_nodes/ComfyUI_Civitai_Gallery/.
4. Restart ComfyUI.

---

更新日志 (2025-09-11)

• 提示词编辑：在“Civitai Images Gallery”中增加了一个“Edit Prompt"的复选框，勾选后开启对图片提示词的编辑能力，方便实时快速的对提示词进行二次创作，对提示词残缺的图片可以进行提示词补充保存等。顺便优化了收藏库图片的加载。

更新日志 (2025-09-07)

• 🎬 视频预览支持：Civitai 图片库现已支持视频浏览。您可以勾选“显示视频”复选框来控制是否显示视频卡。为防止 ComfyUI 界面中自动播放可能导致的崩溃，请在每个图库卡片的右上角找到播放图标 (▶️)。如果出现该图标，您可以将鼠标悬停在卡片上预览视频，或者双击卡片（或点击播放图标）以原始分辨率观看。

更新日志 (2025-09-06)

• 一键加载工作流：图库中包含 ComfyUI 工作流元数据的图片卡现在将持续显示“加载工作流”图标 (🎁)。 点击此图标即可立即将整个工作流加载到当前工作区，就像拖放工作流文件一样。 增强了数据解析的稳定性，以便兼容处理并自动修复来自各种来源的格式错误的 JSON 数据（例如，包含未定义值或 NaN 值），从而提高加载的成功率。

• 模型库与图片库之间的关联：在“Civitai 模型库”节点的模型版本选择窗口中，每个模型版本都添加了一个“🖼️ 查看图片”按钮。 点击此按钮后，“Civitai 图片库”将加载并仅显示该特定模型版本的图片。 在关联模式下，图片库将显示一个清晰的通知栏，指示当前正在查看的模型和版本，并可选择“清除筛选条件”返回正常浏览。

更新日志 (2025-09-05)

• 新节点：Civitai 模型库：添加了一个全新的 Civitai 模型库节点。它允许您在 ComfyUI 中直接从 Civitai 浏览、筛选和下载模型（检查点、LoRA、VAE 等）。

• 模型和资源下载器：已实现适用于所有资源类型的下载器。只需点击新的“已使用资源”查看器或模型库中的“下载”按钮，即可将文件保存到正确的文件夹。这需要一次性设置您的 Civitai API 密钥。

• 高级收藏夹和标签：收藏夹系统已全面改进。您现在可以为收藏的图像添加自定义标签，以便更好地进行整理。

• 增强的 UI 和工作流内存：该节点现在会保存您工作流中的所有 UI 设置（过滤器、选择、排序），并在重新加载时自动恢复。

概述

ComfyUI Civitai Gallery 是一个功能强大的 ComfyUI 自定义节点，它将一个为 Civitai 网站打造的无缝图片浏览器直接集成到了您的工作流中。该节点允许您直接浏览、搜索和选择来自 Civitai 的图片，并能一键将其提示词、负向提示词和原始高分辨率图片导入到您的工作流中。它旨在通过消除在浏览器和 ComfyUI 之间来回切换的需要，从而极大地加速您的创作流程。

本插件的图库拥有一个流畅且响应式的瀑布流（砌体）布局，能够智能地填充所有可用空间，确保了美观高效的浏览体验。

功能特性

• 直连 Civitai 浏览：无需离开 ComfyUI 界面即可浏览来自 Civitai 的图片。
• 高级筛选：可以根据 NSFW 等级、排序方式（最多反应、最新等）、时间范围（天、周等）、特定标签和作者用户名来筛选图片。
• 一键导入：只需单击一张图片即可选中它。当您运行工作流时，该节点将输出：
  • positive_prompt：用于生成该图片的正向提示词。
  • negative_prompt：负向提示词。
  • image：原始的、未经压缩的高分辨率图片。
  • info：一个包含所有其他可用元数据（如采样器、步数、种子、模型哈希等）的详细 JSON 字符串。
• 智能图片加载：只有当 image 输出端口连接到其他节点时，插件才会下载原始图片，从而节省您的时间和带宽。
• 稳定瀑布流布局：一个流畅的、响应式的瀑布流布局，当通过无限滚动加载新图片时，已加载的图片会保持在原位不动。
• 自定义界面：拥有一个自定义样式的、永久可见的滚动条，方便您进行导航。
• 高级灯箱预览器:
  • 双击任意图片即可打开一个全局居中的全屏预览器。
  • 图片查看器: 支持使用鼠标滚轮进行缩放，并通过拖动进行平移。
  • 图库导航: 使用界面上的箭头或键盘方向键（左/右）轻松切换浏览图片。

使用方法

1. 添加节点：在您的 ComfyUI 工作区中按 Tab 键或双击鼠标，搜索 Civitai Gallery，然后将该节点添加到您的图中。
2. 浏览与筛选：
   • 使用节点顶部的下拉菜单和输入框，根据您的需求筛选图片。
   • 如果你是国际用户，可以勾选 "International" 复选框，能得到更好的图片浏览体验。
   • 可以点击 "Refresh" 按钮来应用新的筛选条件，当然一般情况下不需要用到。
   • 在图库区域内向下滚动，即可自动加载更多图片（无限滚动）。
3. 选择图片：在图库中单击任意一张图片卡片。您选中的图片周围会出现一个彩色的边框。
4. 连接输出端口：
   • 将 positive_prompt 和 negative_prompt 输出连接到您的 KSampler 节点或文本显示节点的相应输入上。
   • 如果您需要使用原始图片，请将 image 输出连接到 Preview Image（预览图像）或 Save Image（保存图像）等节点。（请注意：只有当此端口被连接时，插件才会执行下载操作）。
   • 将 info 输出连接到 Show Text（显示文本）节点，以查看所有其他的生成参数。
5. 运行工作流：点击 "Queue Prompt" 执行您的工作流。所选图片的数据将被送入已连接的节点中。

安装说明

1. 导航至您的 ComfyUI 安装目录。
2. 进入 custom_nodes 文件夹。
3. 将此插件的仓库克隆或下载到 custom_nodes 文件夹中。最终的文件夹结构应为 ComfyUI/custom_nodes/ComfyUI_Civitai_Gallery/。
4. 重启 ComfyUI。