Extract CivitAI Model Download Links via:
- Info Files
- Stable Diffusion Directory
- CivitAI Download History
- A handy tool to extract download links for CivitAI Models
- Currently capable of extracting links for:
- LORAs
- Lycoris/ Locon
- Embeddings/ Textual Inversion
- Checkpoints
- Poses
- VAEs
- Aesthetic Gradients
- Upscalers
- Wildcards
- Python v3.4 or higher
- Pip installer
- Chrome v120.x
(Other browsers wont work!)
- only if you want to extract links your your CivitAI Download History.
Tip
This repo works with CivitAIBatchModelDownloader, to batch download CivitAI Models!
- Make sure you have all files required.
D:\CivitAI-DownloadLink-Extractor
│ .gitattributes
│ .gitignore
│ config.ini
│ LICENSE
│ README.md
│ requirements.txt
│
├───CivitAI_Info_Files
│ _InfoFilesHere.txt
│
├───Output
│ _csv files go here.txt
│
└───src
│ civitai_download_history.py
│ info_files.py
│ local_SD.py
│
└───modules
│ civitaihelper.py
│ csv_handler.py
│ process_util.py
└───
- Clone the repo
git clone https://github.com/deepratnaawale/CivitAI-DownloadLink-Extractor.git
- Goto the Downloaded repo
cd CivitAI-DownloadLink-Extractor
- Make virtual enviorenment using pip
(recommended)
python -m venv .venv
-
Activate Virtual Enviorenment
- For Windows
.venv\Scripts\activate.bat
- For Linux/ MacOS
.venv\Scripts\activate.sh
-
Install Requirements
pip install -r requirements.txt
- Run program!
python extract_links.py
- Copy all ".info" files to CivitAI_Info_Files directory. You do not need to seperate the info files based on model type, CSVs will do that for you!
Note
ALL PATHS IN CONFIG.INI
MUST BE ABSOLUTE PATHS.
- Edit
config.ini
.
[SD-DIR]
dir = <SD Folder Location>
[Output] # This is optional, Default is Project Directory/Output/
dir = <Custom Output Dir>
You NEED Chrome(>=v120.x) for this. Once you start this option, DO NOT touch your keyboard or mouse.
Please be patient!
-
You are required to be logged into CivitAI in chrome (the program will redirect you to login page if you arent.)
-
The script will automatically try to launch chrome from default paths.
- If path is not found then it will ask you to open chrome.
- The program will get the path from the executable/ file for any OS.
-
The browser will be closed and started in debug mode, so we can control it with program.
-
Once the Debug Browser starts (You cant really tell it's in debug mode)
- The script will automatically navigate to user account (to check if you are logged in).
- If you are logged in it will navigate to your download history automatically.
- It will scroll down to the bottom of the page (PLEASE DO NOT minimize, mouse click, or use the keyboard; the script WILL FAIL! Unfortunately, I couldn't find any way around this).
- All your CSVs are with the following format:
Sr.no | Model ID | Model Name | Model URL |
---|---|---|---|
1 | 4201 | Realistic Vision V4.0 | https://civitai.com/api/download/models/114367 |
2 | ... | ... | ... |
3 | ... | ... | ... |
- Qyabghuyn94 for civitai-helper.py: https://github.com/quanghuyn94/anything-model-batch-downloader/
- I made this script to use with Runpod.
- If you are looking to batch download these extracted models to Runpod, goto my CivitAIBatchModelDownloader repository. It uses multi threading to dramatically increase download speeds! You don't even need to wait for everything to download!