ComfyUI Browser Load Image
Enhanced Load Image/Video nodes for ComfyUI with visual thumbnail browser, hover preview, draggable/resizable/maximizable modal
Nodes (3)
ComfyUI Browser Load Image
Enhanced Load Image / Video nodes for ComfyUI with a visual browser interface.

✨ Features
- Visual Media Browser — Browse and select images/videos from the input folder with a thumbnail grid modal
- Hover Preview — Hover over any thumbnail to see a large preview with filename and original dimensions
- Configurable Preview Size — Choose preview size from dropdown (320 / 480 / 640 / 768 / 1024)
- Draggable Modal — Drag the title bar to reposition the modal
- Resizable Modal — Drag the bottom-right corner to resize (min 400×300)
- Maximize Button — Click □ to toggle fullscreen
- Remember Layout — Check "记住窗口" to persist your preferred window position and size (also remembers the last folder)
- Search & Filter — Real-time search to quickly find files by name (searches all folders)
- Keyboard Navigation — Arrow keys, PageUp/Down, Home/End for scrolling; ESC to close
- Subfolders (one level) — Group media into folders under
input/; navigate with folder cards and a breadcrumb - File Management — Delete, move, rename and batch-select files and folders right in the modal
- Trash — Deletions are recoverable: restore to the original location, or purge permanently
📦 Included Nodes
| Node | Description | Output | |------|-------------|--------| | Browser Load Image | Load images with visual browser | IMAGE, MASK | | Browser Load Video to Image | Convert video to image sequence | IMAGE, INT, INT | | Browser Load Video | Load video (standard VIDEO type) | VIDEO |
🔧 Installation
Option 1: ComfyUI Manager (Recommended)
Search for ComfyUI_BrowserLoadImage in ComfyUI Manager and install.
Option 2: Manual Installation
cd ComfyUI/custom_nodes
git clone https://github.com/flashcol/ComfyUI_BrowserLoadImage.git
cd ComfyUI_BrowserLoadImage
pip install -r requirements.txt
Restart ComfyUI after installation.
Dependencies
- Image node: No extra dependencies (uses PIL/Pillow included with ComfyUI)
- Video nodes: Requires
opencv-python(pip install opencv-python)
🖱️ Usage
- Add a Browser Load Image node to your workflow
- Click the "浏览图片" (Browse Images) button
- Browse, search, and click a thumbnail to select it; click a folder card to enter it
- The selected file is set as the node's input (subfolder files look like
sub/a.png)
🗂️ File Management
All operations act on the input/ directory and are available in the browser modal.
- Subfolders — One level only. Click a folder card to enter it, use the breadcrumb to go back. Create folders with 新建文件夹.
- Delete — Click ✕ on a card, or turn on 管理 (manage mode), multi-select, and press
Delete. Files and whole folders are moved toinput/.trash/, never deleted permanently. - Move — In manage mode, select files and click 移动到…. On a name clash a
(1)suffix is added; nothing is overwritten. - Rename — Click ✎ on a card, or select one item in manage mode and press
F2. A file's extension cannot be changed.
Moving or renaming changes file paths. Nodes in other saved workflows that reference the old path will not find the file.
Trash
- Open it with 回收站 (N) in the toolbar (hover to see the size it occupies). The trash is shared by all nodes: it always shows every deleted image, video and folder, with an optional 全部 / 图片 / 视频 filter. Each deletion records the original path in
input/.trash/.index.json. - Restore — Back to the original path. A missing original folder is recreated; on a name clash a
(n)suffix is added. Entries without an index record are restored to the root. - Inspect folders — Click 👁 on a folder card to look inside a deleted folder. There you can restore or purge individual files; once a folder entry is emptied it disappears from the trash.
- Purge / Empty trash — Irreversible, with a confirmation dialog. The trash is never cleaned automatically.
HTTP API
Registered on ComfyUI's server (all paths are validated to stay inside input/):
GET /bli/list, POST /bli/delete, /bli/move, /bli/mkdir, /bli/rename,
GET /bli/trash/list, POST /bli/trash/restore, /bli/trash/purge.
📁 Project Structure
ComfyUI_BrowserLoadImage/
├── __init__.py # Plugin entry point (registers nodes and API)
├── api/
│ └── routes.py # /bli/* HTTP endpoints
├── nodes/
│ ├── utils.py # Shared utilities (media listing, path resolving)
│ ├── fileops.py # Delete / move / rename / trash operations
│ ├── browser_load_image.py
│ ├── browser_load_video.py
│ └── browser_load_video_standard.py
├── web/
│ ├── js/browserLoadImage.js # Frontend extension
│ └── css/browserLoadImage.css # Stylesheet reference
├── config/settings.py # Constants
└── tests/
├── test_nodes.py # Node and utility tests
└── test_fileops.py # File management and trash tests
🧪 Running Tests
cd ComfyUI/custom_nodes/ComfyUI_BrowserLoadImage
python -m pytest tests/ -v