ComfyUI LoRA Manager: Revolutionize Your Workflow

Introduction

In the world of AI and machine learning, managing LoRA (Low-Rank Adaptation) models can be a challenging task. However, the ComfyUI LoRA Manager comes to the rescue as a comprehensive toolset designed to streamline the organization, downloading, and application of LoRA models in ComfyUI. This blog post will delve deep into the features, installation process, usage, and more aspects of the ComfyUI LoRA Manager, providing you with all the information you need to make the most of this powerful tool.

Overview of ComfyUI LoRA Manager

The ComfyUI LoRA Manager is an ultimate companion for ComfyUI users. It offers a range of features that make working with LoRA models faster, smoother, and significantly easier. With powerful capabilities like recipe management, checkpoint organization, and one-click workflow integration, it has become an essential tool in the AI community.

Key Features

Let’s take a closer look at some of the key features that set the ComfyUI LoRA Manager apart:

High Performance

Fast model loading and browsing: The manager ensures that you can quickly access and view your LoRA models without any significant delays. Whether you have a small collection or a large one, the loading times are optimized for efficiency.
Smooth scrolling through large collections: When dealing with a vast number of models, scrolling through them can be a hassle. But the ComfyUI LoRA Manager provides a smooth scrolling experience, allowing you to navigate through your collection effortlessly.

Rich Model Integration

Direct download from CivitAI: You can directly download LoRA models from CivitAI, a popular platform for AI models. This feature saves you the time and effort of searching for models elsewhere and then manually transferring them to your ComfyUI environment.
Preview images and videos: Before downloading a model, you can preview its associated images and videos. This gives you a better understanding of what the model can do and helps you make an informed decision about whether to use it or not.
Model descriptions and version selection: The manager provides detailed descriptions of each model, along with the option to select different versions. This allows you to choose the most suitable version of a model based on your specific requirements.
Trigger words at a glance: Trigger words are essential for using LoRA models effectively. The ComfyUI LoRA Manager displays these trigger words at a glance, making it easy for you to incorporate them into your workflows.
One-click workflow integration with preset values: With just one click, you can integrate a LoRA model into your ComfyUI workflow. The manager also allows you to set preset values, which further simplifies the process and saves you time.

Checkpoint Management

Scan and organize checkpoint models: The manager can scan your system for checkpoint models and organize them in a structured manner. This makes it easier for you to keep track of your models and find the ones you need quickly.
Filter and search your collection: You can use filters and search functionality to narrow down your model collection. Whether you’re looking for a specific model by name, type, or other criteria, the manager makes it easy to find what you’re looking for.
View and edit metadata: Metadata provides important information about your models, such as their creation date, author, and description. The ComfyUI LoRA Manager allows you to view and edit this metadata, ensuring that your model collection is well-documented.
Clean up and manage disk space: Over time, your model collection can take up a significant amount of disk space. The manager helps you clean up your collection by identifying and removing duplicate models, freeing up valuable disk space.

LoRA Recipes

Save and share favorite LoRA combinations: You can save your favorite LoRA combinations as recipes. This allows you to reuse these combinations in the future without having to recreate them from scratch. You can also share these recipes with other users in the community.
Preserve generation parameters for future reference: When generating images or other outputs using LoRA models, you can preserve the generation parameters. This is useful for reproducing the same results in the future or for sharing the parameters with others.
Quick application to workflows: Recipes can be quickly applied to your ComfyUI workflows. This saves you time and effort, especially when working on multiple projects.
Import/export functionality for community sharing: The manager supports the import and export of recipes, which facilitates community sharing. You can download recipes created by other users and contribute your own recipes to the community.

User Friendly

One-click access from ComfyUI menu: You can access the LoRA manager directly from the ComfyUI menu with just one click. This makes it convenient to use and saves you time navigating through different interfaces.
Context menu for quick actions: The manager provides a context menu that allows you to perform quick actions on your models, such as copying, deleting, or viewing details. This enhances the user experience and makes it easier to manage your model collection.
Custom notes and usage tips: You can add custom notes and usage tips to your models. This helps you remember important information about the models and how to use them effectively.
Multi-folder support: The manager supports multiple folders, allowing you to organize your models in a more flexible way. You can create different folders for different types of models or projects.
Visual progress indicators during initialization: When initializing the manager, visual progress indicators are provided. This gives you an idea of how long the initialization process will take and helps you plan your work accordingly.

Release Notes

The ComfyUI LoRA Manager has been continuously updated with new features and improvements. Let’s take a look at some of the recent release notes:

v0.8.20

LM Civitai Extension: A browser extension has been released through the Chrome Web Store. This extension works seamlessly with the LoRA Manager to enhance the Civitai browsing experience. It shows which models are already in your local library, enables one-click downloads, and provides queue and parallel download support.
Enhanced Lora Loader: Support for nunchaku has been added, improving convenience when working with ComfyUI-nunchaku workflows. New template workflows are also available for quick onboarding.
WanVideo Integration: A WanVideo Lora Select (LoraManager) node has been introduced, which is compatible with ComfyUI-WanVideoWrapper. This enables streamlined lora usage in video workflows, and a template workflow is provided to help you get started quickly.

v0.8.19

Analytics Dashboard: A new Statistics page has been added, providing comprehensive visual analysis of model collection and usage patterns. This helps you gain better insights into your library.
Target Node Selection: Workflow integration has been enhanced with intelligent target choosing when sending LoRAs/recipes to workflows with multiple loader/stacker nodes. A visual selector now appears, showing node color, type, ID, and title for precise targeting.
Enhanced NSFW Controls: Support for setting NSFW levels on recipes has been added, with automatic content blurring based on user preferences.
Customizable Card Display: New display settings allow users to choose whether card information and action buttons are always visible or only revealed on hover.
Expanded Compatibility: Support for efficiency-nodes-comfyui has been added in the Save Recipe and Save Image nodes, and compatibility with ComfyUI_Custom_Nodes_AlekPet has been fixed.

v0.8.18

Custom Example Images: The ability to import your own example images for LoRAs and checkpoints has been added, with automatic metadata extraction from embedded information.
Enhanced Example Management: New action buttons have been added to set specific examples as previews or delete custom examples.
Improved Duplicate Detection: The “Find Duplicates” feature has been enhanced with a hash verification feature to eliminate false positives when identifying duplicate models.
Tag Management: Tag editing functionality has been added, allowing users to customize and manage model tags.
Advanced Selection Controls: The Ctrl+A shortcut has been implemented for quickly selecting all filtered LoRAs, automatically entering bulk mode when needed.
Note: Cache file functionality has been temporarily disabled pending rework.

v0.8.17

Duplicate Model Detection: A “Find Duplicates” functionality has been added for LoRAs and checkpoints using model file hash detection. This enables convenient viewing and batch deletion of duplicate models.
Enhanced URL Recipe Imports: The import recipe via URL functionality has been optimized using CivitAI API calls instead of web scraping. It now supports all rated images (including NSFW) for recipe imports.
Improved TriggerWord Control: The TriggerWord Toggle node has been enhanced with a new default_active switch to set the initial state (active/inactive) when trigger words are added.
Centralized Example Management: A “Migrate Existing Example Images” feature has been added to consolidate downloaded example images from model folders into central storage with customizable naming patterns.
Intelligent Word Suggestions: Smart trigger word suggestions have been implemented by reading class tokens and tag frequency from safetensors files. Recommendations are displayed when editing trigger words.
Model Version Management: A “Re-link to CivitAI” context menu option has been added for connecting models to different CivitAI versions when needed.

v0.8.16

Dramatic Startup Speed Improvement: A cache serialization mechanism has been added for significantly faster loading times, especially beneficial for large model collections.
Enhanced Refresh Options: The functionality has been extended with a “Full Rebuild (complete)” option alongside “Quick Refresh (incremental)” to fix potential memory cache issues without requiring application restart.
Customizable Display Density: The compact mode has been replaced with adjustable display density settings for personalized layout customization.
Model Creator Information: Creator details have been added to model information panels for better attribution.
Improved WebP Support: The Save Image node has been enhanced with workflow embedding capability for WebP format images.
Direct Example Access: A “Open Example Images Folder” button has been added to card interfaces for convenient browsing of downloaded model examples.
Enhanced Compatibility: Full ComfyUI Desktop support has been added for the “Send lora or recipe to workflow” functionality.
Cache Management: Settings have been added to clear existing cache files when needed.
Bug Fixes & Stability: Various improvements have been made for overall reliability and performance.

v0.8.15

Enhanced One-Click Integration: The copy button has been replaced with a direct send button, allowing LoRAs/recipes to be sent directly to your current ComfyUI workflow without needing to paste.
Flexible Workflow Integration: You can click to append LoRAs/recipes to existing loader nodes or Shift+click to replace content. Additional right-click menu options for “Send to Workflow (Append)” or “Send to Workflow (Replace)” are also available.
Improved LoRA Loader Controls: Header drag functionality has been added for proportional strength adjustment of all LoRAs simultaneously (including CLIP strengths when expanded).
Keyboard Navigation Support: Page Up/Down for page scrolling, Home key to jump to top, and End key to jump to bottom have been implemented for faster browsing through large collections.

v0.8.14

Virtualized Scrolling: The rendering mechanism has been completely rebuilt for smooth browsing with no lag or freezing. It now supports virtually unlimited model collections with optimized layouts for large displays, improving space utilization and user experience.
Compact Display Mode: A space-efficient view option has been added that displays more cards per row (7 on 1080p, 8 on 2K, 10 on 4K).
Enhanced LoRA Node Functionality: Comprehensive improvements have been made to LoRA loader/stacker nodes, including real-time trigger word updates (reflecting any change anywhere in the LoRA chain for precise updates) and an expanded context menu with “Copy Notes” and “Copy Trigger Words” options for faster workflow.

v0.8.13

Enhanced Recipe Management: A “Find duplicates” feature has been added to identify and batch delete duplicate recipes, with duplicate detection notifications during imports.
Improved Source Tracking: Source URLs are now saved with recipes imported via URL, allowing users to view original content with one click or manually edit links.
Advanced LoRA Control: Double-clicking LoRAs in Loader/Stacker nodes allows access to expanded CLIP strength controls for more precise adjustments of model and CLIP strength separately.
Lycoris Model Support: Compatibility with Lycoris models has been added for expanded creative options.
Bug Fixes & UX Improvements: Various issues have been resolved, and the overall user experience has been enhanced with numerous optimizations.

v0.8.12

Enhanced Model Discovery: An alphabetical navigation bar has been added to the LoRAs page for faster browsing through large collections.
Optimized Example Images: The download logic has been improved to automatically refresh stale metadata before fetching example images.
Model Exclusion System: A new right-click option has been added to exclude specific LoRAs or checkpoints from management.
Improved Showcase Experience: Interaction in the LoRA and checkpoint showcase areas has been enhanced for better usability.

v0.8.11

Offline Image Support: Functionality has been added to download and save all model example images locally, ensuring access even when offline or if images are removed from CivitAI or the site is down.
Resilient Download System: A pause/resume capability with checkpoint recovery has been implemented, which persists through restarts or unexpected exits.
Bug Fixes & Stability: Various issues have been resolved to enhance overall reliability and performance.

v0.8.10

Standalone Mode: The LoRA Manager can now be run independently from ComfyUI for a lightweight experience that works even with other stable diffusion interfaces.
Portable Edition: A new one-click portable version has been introduced for easy startup and updates in standalone mode.
Enhanced Metadata Collection: Support for the SamplerCustomAdvanced node has been added in the metadata collector module.
Improved UI Organization: The Lora Loader node height has been optimized to display up to 5 LoRAs at once with scrolling capability for larger collections.

Installation

There are several ways to install the ComfyUI LoRA Manager. Let’s explore each option in detail:

Option 1: ComfyUI Manager (Recommended for ComfyUI users)

Open ComfyUI.
Go to Manager > Custom Node Manager.
Search for lora-manager.
Click Install.

This option is the simplest and most straightforward way to install the LoRA Manager if you are already using ComfyUI. The ComfyUI Manager takes care of all the installation steps for you.

Option 2: Portable Standalone Edition (No ComfyUI required)

Download the Portable Package.
Copy the provided settings.json.example file to create a new file named settings.json in the comfyui-lora-manager folder.
Edit settings.json to include your correct model folder paths and CivitAI API key.
Run run.bat.

This option is suitable for users who do not want to use ComfyUI or want to run the LoRA Manager independently. The portable standalone edition provides a lightweight option for managing your model and recipe collection.

Option 3: Manual Installation

git clone https://github.com/willmiao/ComfyUI-Lora-Manager.git
cd ComfyUI-Lora-Manager
pip install -r requirements.txt

If you prefer to have more control over the installation process, you can choose the manual installation option. This involves cloning the repository from GitHub and installing the required dependencies using pip.

Usage

Once you have installed the ComfyUI LoRA Manager, you can start using it to manage your LoRA models. Here’s how:

Accessing the LoRA Manager

There are two ways to access the LoRA manager:

Click the “Launch LoRA Manager” button in the ComfyUI menu.
Visit http://localhost:8188/loras directly.

Using the LoRA Manager Interface

From the interface, you can perform the following actions:

Browse and organize your LoRA models: You can view all your LoRA models in a structured manner and organize them into folders or categories as needed.
Download models directly from CivitAI: If you find a model on CivitAI that you want to use, you can download it directly from the LoRA manager interface.
Automatically fetch or manually set preview images: The manager can automatically fetch preview images for your models, or you can manually set them if needed.
View and copy trigger words associated with each LoRA: Trigger words are important for using LoRA models effectively. You can view and copy these trigger words from the manager interface.
Add personal notes and usage tips: You can add your own personal notes and usage tips to each model, which can be helpful for future reference.

Using LoRAs in Your Workflow

To use LoRAs in your workflow, follow these steps:

Add the “Lora Loader (LoraManager)” node to your workflow.
Select a LoRA in the manager interface.
Click the copy button or use the right-click menu “Copy LoRA syntax”.
Paste into the Lora Loader node’s text input.
The node will automatically apply preset strength and trigger words.

Filename Format Patterns for Save Image Node

The Save Image Node supports dynamic filename generation using pattern codes. You can customize how your images are named using the following format patterns:

Available Pattern Codes

%seed% – Inserts the generation seed number
%width% – Inserts the image width
%height% – Inserts the image height
%pprompt:N% – Inserts the positive prompt (limited to N characters)
%nprompt:N% – Inserts the negative prompt (limited to N characters)
%model:N% – Inserts the model/checkpoint name (limited to N characters)
%date% – Inserts current date/time as “yyyyMMddhhmmss”
%date:FORMAT% – Inserts date using custom format with:
- yyyy – 4-digit year
- yy – 2-digit year
- MM – 2-digit month
- dd – 2-digit day
- hh – 2-digit hour
- mm – 2-digit minute
- ss – 2-digit second

Examples

image_%seed% → image_1234567890
gen_%width%x%height% → gen_512x768
%model:10%_%seed% → dreamshape_1234567890
%date:yyyy-MM-dd% → 2025-04-28
%pprompt:20%_%seed% → beautiful landscape_1234567890
%model%_%date:yyMMdd%_%seed% → dreamshaper_v8_250428_1234567890

You can combine multiple patterns to create detailed, organized filenames for your generated images.

Standalone Mode

You can now run the LoRA Manager independently from ComfyUI:

For ComfyUI users

Launch ComfyUI with LoRA Manager at least once to initialize the necessary path information in the settings.json file.
Make sure dependencies are installed: pip install -r requirements.txt.

From your ComfyUI root directory, run:

python custom_nodes\comfyui-lora-manager\standalone.py

Access the interface at: http://localhost:8188/loras.

You can specify a different host or port with arguments:

python custom_nodes\comfyui-lora-manager\standalone.py --host 127.0.0.1 --port 9000

For non-ComfyUI users

Copy the provided settings.json.example file to create a new file named settings.json.
Edit settings.json to include your correct model folder paths and CivitAI API key.
Install required dependencies: pip install -r requirements.txt.
Run standalone mode:
```
python standalone.py
```
Access the interface through your browser at: http://localhost:8188/loras.

This standalone mode provides a lightweight option for managing your model and recipe collection without needing to run the full ComfyUI environment, making it useful even for users who primarily use other stable diffusion interfaces.

Contributing

The developers of the ComfyUI LoRA Manager appreciate your interest in contributing to the project. However, as the project is currently in its early stages and undergoing rapid development and refactoring, they are temporarily not accepting pull requests.

But your feedback and ideas are extremely valuable. You can:

Open issues for any bugs you encounter.
Submit feature requests through GitHub issues.
Share your suggestions for improvements.

The developers look forward to potentially accepting code contributions once the project architecture stabilizes.

Credits

The ComfyUI LoRA Manager has been inspired by and benefited from other excellent ComfyUI extensions:

ComfyUI-SaveImageWithMetaData – For the image metadata functionality.
rgthree-comfy – For the lora loader functionality.

Support

If you find this project helpful, consider supporting its development:

ko-fi
Patreon
WeChat: Click to view QR code

Community

Join the Discord community for support, discussions, and updates:
Discord Server

In conclusion, the ComfyUI LoRA Manager is a powerful and user-friendly tool that simplifies the management of LoRA models in ComfyUI. With its rich features, easy installation process, and flexible usage options, it is a must-have for anyone working with LoRA models. Whether you are a beginner or an experienced user, the ComfyUI LoRA Manager can help you streamline your workflow and achieve better results. So why not give it a try and see the difference it can make?

ComfyUI LoRA Manager: Master AI Model Management & Workflow Optimization