The Simplest FLUX Beginner's Tutorial – One-Click Setup

Everyone is talking about FLUX, and it works great. But for beginners, the first step of downloading and choosing a model can be confusing. There are so many models—how do you know which one to download?

The number of nodes is also overwhelming for beginners.

Today, I’ll show you the simplest FLUX beginner process.

I recommend the following three versions. Choose one based on your needs:

• DEV Integrated Version, file size: 16GB
• Schnell Integrated Accelerated Version, file size: 16GB
• NF4 Integrated VRAM Version, file size: 11GB

Model installation location: ComfyUI\models\checkpoints

Each version includes CLIP and EVA integration, so you only need to download one large model. The usage is the same as for 1.5 and XL models.

The DEV integrated version has the best performance, Schnell is the fastest, and NF4 has performance close to DEV with the lowest GPU requirements. If your GPU is strong, go with DEV; if not, go with NF4.

You can easily set up the FLUX workflow in one step by using the default workflow.

Select either DEV 16GB or Schnell 16GB as the large model. These two versions support the default loader.

Search for FLUX.

Add the FLUX text encoder, replacing the original positive prompt, as shown in the link below. Leave the negative prompt blank.

At this point, we’ve completed our FLUX workflow. Isn’t this the simplest workflow you’ve seen?

Enter the prompt:

Creative works, including apples, oranges, strawberries, forming little mice.

You can input the same prompt in both positive prompt boxes.

The default guidance is set to 3.5, which can remain unchanged.

Set the image size to a custom dimension, such as 800x1200.

You can keep the steps at 20 in the sampler, though in the screenshot it’s adjusted to 16 for speed.

Set CFG to 1. Everything else can stay unchanged.

The results are exactly the same as with the 11GB model + CLIP + VAE version. This integration makes it more convenient and simpler.

Change the prompt:

A photo of a rainbow-colored caterpillar on a moss-covered branch, backlit, translucent, glowing, hyper-realistic, with soft natural rim light and cinematic lighting, concept art by Arti Stincent.

This captivating image, taken with a Canon EOS R5 and a Canon RF 85mm f/1.2 lens, beautifully portrays a girl immersed in a floral wonderland. The unique perspective, shot from above, showcases the girl surrounded by a vibrant array of flowers, their natural textures, and the soft sunlight filtering through. The warm light illuminates her face, creating a radiant glow that accentuates her features. The sharp lens captures the delicate details of the flowers and the girl’s expression, drawing the viewer into the moment. It’s a whimsical and dream-like scene, inviting viewers to imagine the scent of the flowers and the peacefulness of being surrounded by nature’s beauty.

Schnell Model Workflow

The Schnell model workflow is exactly the same as the DEV version. We only need to replace the model and change the sampling steps to 6–10.

NF4 Model Workflow

The NF4 model requires a specific loader. You need to install the N4F node.

Once installed, search for NF4.

Choose the NF4 model and link it to the corresponding model. This completes the NF4 model workflow.

Using LORA Models

For example, use the following LORA:

Prompt:

The photo was taken with Canon EOS R5 and Canon RF 85mm f/1.2 lenses, depicting a man wearing armor and sporting a beard, immersed in flames. Wukong is wearing armor, with a golden cudgel behind him. The unique perspective captured from above shows Wukong surrounded by vibrant flame effects in a sea of flowers, with close-ups of his facial features.

Workflow Download

Shared file via cloud: 0904 FLUX

Link: https://pan.baidu.com/s/1y9wzBDHZzNnL6PUQoq2z4A?pwd=8787
Password: 8787