How to use Flux LoRA on ComfyUI

Updated Categorized as Tutorial Tagged , , 2 Comments on How to use Flux LoRA on ComfyUI
All LoRAs enabled

Flux is a state-of-the-art image model. It excels in generating realistic photos and following the prompt, but some styles can be challenging to achieve. In addition to crafting the prompt, supplementing the Flux base model with LoRA models is another valuable trick in your toolbox.

In this article, you will learn how to use up to 3 LoRA models in ComfyUI.

Software

We will use ComfyUI, an alternative to AUTOMATIC1111. You can use it on Windows, Mac, or Google Colab. If you prefer using a ComfyUI service, Think Diffusion offers our readers an extra 20% credit.

Read the ComfyUI beginner’s guide if you are new to ComfyUI. See the Quick Start Guide if you are new to AI images and videos.

Take the ComfyUI course to learn how to use ComfyUI step-by-step.

Using LoRA on ComfyUI

Step 0: Update ComfyUI

The easiest way to update ComfyUI is through the ComfyUI Manager. Click Manager > Update ComfyUI.

Restart ComfyUI and reload the page.

Step 1: Load the ComfyUI workflow

Download the Flux LoRA workflow JSON file below.

Drop it to your ComfyUI.

This workflow uses up to 3 LoRA models but with the third one disabled. I will show you how to enable/disable any one of them.

Step 2: Download the Flux model

Download the Flux1 dev FP8 checkpoint.

Put the model file in the folder ComfyUI > models > checkpoints.

(If you use my ComfyUI Colab notebook, simply select the Flux1_dev model.)

Refresh the ComfyUI page and select the flux model in the Load Checkpoint dropdown menu.

Step 3: Download the Flux LoRA models

Download the following LoRA models

Put them in the folder  ComfyUI > models > loras.

(If you use my ComfyUI Colab notebook, put them in your Google Drive folder AI_PICS > models > Lora. Restart the notebook to take effect.)

Select them in the dropdown menus of the Load LoRA nodes.

Step 4: Generate an image.

Press Queue Prompt to generate an image.

Enable/Disable a LoRA

To enable or disable a LoRA model, right-click on the Load LoRA node and select Bypass.

The node will be grayed out.

Keyboard shortcut: Select the node and press Ctrl + B to enable or disable bypass.

Compare the images with and without the LoRAs:

Avatar

By Andrew

Andrew is an experienced engineer with a specialization in Machine Learning and Artificial Intelligence. He is passionate about programming, art, photography, and education. He has a Ph.D. in engineering.

2 comments

  1. Excellent walk through.
    It run really slow though, when I looked at Task Manager & resouce monitor, the CPU was 1% and two (of 24) CPUs where being used heavily.
    Is there some reason its not using GPU?

Leave a comment

Your email address will not be published. Required fields are marked *