Model Downloader for ComfyUI

<div align="center"> <a href="https://www.python.org/"> <img src="https://img.shields.io/badge/Python-3.10%2B-3776AB?logo=python&logoColor=white" alt="Python 3.10+"> </a> <a href="https://github.com/comfyanonymous/ComfyUI"> <img src="https://img.shields.io/badge/ComfyUI-Custom%20Node-5C5CDB" alt="ComfyUI custom node"> </a> <a href="LICENSE"> <img src="https://img.shields.io/badge/License-AGPL--3.0-blue" alt="AGPL-3.0 license"> </a> <img src="https://img.shields.io/badge/Version-0.3.0-informational" alt="Version 0.3.0"> </div> <div align="center"> <picture> <source media="(prefers-color-scheme: light)" srcset="https://github.com/ciri/comfyui-model-downloader/blob/main/assets/logo.svg?raw=true"> </picture> </div>

Introduction

This project provides a model downloader node for ComfyUI, designed to simplify the process of downloading and managing models in environments with restricted access or complex setup requirements. It aims to enhance the flexibility and usability of ComfyUI by enabling seamless integration and management of machine learning models.

Features

• Easy Model Downloading: Simplify the process of downloading models directly within the ComfyUI environment.
• Repositories: Currently supports hugging face and CivitAI.
• User-friendly: Designed with a focus on ease of use, making model management accessible to users of all skill levels.

Available Nodes

HF Downloader

Parameters:

• repo_id: Hugging Face repo ID
• filename: filename to download from the repository
• save_dir: destination directory
• overwrite: overwrite existing file if it exists
• hf_token: optional Hugging Face access token for gated or private repositories
• filename output: downloaded filename, for connecting to string-compatible nodes

CivitAI Downloader

Parameters:

• model_id: CivitAI model ID
• version_id: CivitAI model version ID
• api_key: optional CivitAI API key
• save_dir: destination directory
• filename output: downloaded filename, for connecting to string-compatible nodes

Tiny test workflows

Load the credential-free workflows in examples/workflows to test Hugging Face downloads, CivitAI downloads, the Auto Model Finder, and a complete checkpoint-to-image flow. See examples/workflows/README.md for download sizes, usage, and cleanup instructions.

Auto Model Finder (Experimental)

Scans the current workflow for missing known model files and looks for matching Hugging Face repositories. Connect its repo_id, filename, and local_path outputs directly to the HF Downloader.

When multiple models are missing, choose one from select_model. Choices include the ComfyUI model directory (for example, checkpoints/model.safetensors) so same-named files remain distinct. Click Rescan models after changing the canvas to update this list without running the workflow. The scan also runs on every queue; if a previously selected model is no longer missing, the node safely uses the first current match instead.

Installation

Via ComfyUI Manager (Recommended)

1. Open ComfyUI Manager
2. Select Custom Nodes Manager
3. Search for "ComfyUI Model Downloader"
4. Install

Via Git Clone

Clone the repository or download the extension directly into your ComfyUI project's custom_nodes folder:

git clone https://github.com/ciri/comfyui-model-downloader.git

Usage

To use the model downloader within your ComfyUI environment:

1. Open your ComfyUI project.
2. Find the HF Downloader or CivitAI Downloader node.
3. Configure the node properties with the URL or identifier of the model you wish to download and specify the destination path.
4. Execute the node to start the download process.
5. To avoid repeated downloading, make sure to bypass the node after you've downloaded a model.

Download destinations follow ComfyUI's model folder configuration, including directories marked as default in extra_model_paths.yaml.

Roadmap (tentative)

• [x] Add persistence for auto model finder between runs
• [ ] Add more model finders (including CivitAI)
• [ ] Add more downloaders
• [x] Add authentication for HF Downloader

Contributing

Contributions are welcome! Please:

• Fork the repository
• Create a feature branch
• Submit a pull request

Testing

Install the runtime dependencies and run the complete test suite:

python -m pip install -r requirements.txt
python -m unittest discover -s tests -v

Support

For support, questions, or contributions, please open an issue on the GitHub repository page. Contributions are welcome!

License

GNU Affero General Public License v3.0