Jump to content

AI Art Generation Handbook/FLUX/ComfyUI

From Wikibooks, open books for an open world

Comfy UI

[edit | edit source]

Click here to see how to install ComfyUI first.


Depending on your GPU's VRAM capacity, it can be recommended to use

FLUX.1-schnell (recommended if your GPU's VRAM requirement is less than 12 GB VRam / want faster image generation)

FLUX.1-dev (recommended if your GPU's VRAM requirement is more than 12 GB VRam)


Then, download the following files as followed and paste it into the related folders :

File Name / Link Installed Folders Location
FLUX.1-dev

flux1-dev.safetensors https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/flux1-dev.safetensors OR FLUX.1-schnell flux1-schnell.safetensors https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/main/flux1-schnell.safetensors

...\ComfyUI_windows_portable\ComfyUI\models\unet
FLUX.1-dev

diffusion_pytorch_model.safetensors https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/vae/diffusion_pytorch_model.safetensors OR FLUX.1-schnell diffusion_pytorch_model.safetensors https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/main/vae/diffusion_pytorch_model.safetensors

..\ComfyUI_windows_portable\ComfyUI\models\vae
clip_l.safetensors

t5xxl_fp16.safetensors

t5xxl_fp8_e4m3fn.safetensors

https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main

..\ComfyUI_windows_portable\ComfyUI\models\clip

And then, you can download the workflow files json for FLUX here