ComfyUI Multi-Image Loader is a specialized node extension that enables seamless, interactive batch image loading directly within the ComfyUI interface. Say goodbye to manual folder paths and file-system navigation—just drag, drop, and process.
- Interactive Drag & Drop: Upload multiple images directly into the node widget from your local machine.
- No Folder Required: Bypasses the need for pre-defined input folders; images are handled via a dedicated internal upload route.
- Unified Batch Tensor: Automatically combines all uploaded images into a single
IMAGEbatch tensor[B, H, W, C]. - Smart Resizing: Includes a
resize_to_firstmode to automatically harmonize inconsistent image dimensions based on the first loaded file. - Persistent Storage: The list of uploaded images is saved directly in your ComfyUI workflow JSON for easy multi-session work.
- Visual Previews: (Coming Soon/Frontend) Interactive thumbnail previews for managed assets.
- MultiImageLoader Node: The core backend node responsible for tensor stacking and resizing.
- JavaScript Extension: Injects a custom drag-and-drop / file-picker widget into the ComfyUI graph.
- Navigate to your
ComfyUI/custom_nodes/folder. - Clone the repository:
git clone https://github.com/Latentnaut/ComfyUI-Multi-Image-Loader.git
- Restart ComfyUI and refresh your browser.
- Add the Load Multiple Images 🖼️ node (found under
image/loaders). - Click the widget or drag multiple files directly onto the node.
- Choose your
resize_mode:resize_to_first: Resizes all images to match the dimensions of the first one in the batch.none: Attempt to stack (will fail if dimensions differ).
- Connect the output
image_batchto any node that accepts batch image tensors (e.g., VAE Encode, Upscalers).
Feedback and contributions are welcome! If you encounter issues or have feature requests for the UI, please open an issue on the repository.
Created by Latentnaut