ComfyUI - How to install FLUX - FLUX running on GPUs with only 4GB.

In this video, I demonstrate three different methods for installing and using FLUX in ComfyUI. These methods cater to high-end GPUs as well as entry-level GPUs with only 4GB of VRAM.


First method links:
Clips: https://huggingface.co/comfyanonymous...

VAE: https://huggingface.co/black-forest-l...

Flux1-dev: https://huggingface.co/black-forest-l...
Flux1-schnell: https://huggingface.co/black-forest-l...

FLUX1-Dev workflow(just click the blue button): https://comfyui-wiki.com/tutorial/adv...

FLUX1-Schnell workflow(just click the blue button): https://comfyui-wiki.com/tutorial/adv...

FLUX1 Low VRAM workflow(just click the blue button): https://comfyui-wiki.com/tutorial/adv...



Second method links:
flux1-dev-fp8(16+GB VRAM): https://huggingface.co/Comfy-Org/flux...
flux1-schnell-fp8(8+GB VRAM): https://huggingface.co/Comfy-Org/flux...



Third method links:
Flux model GGUF Dev: https://huggingface.co/city96/FLUX.1-...
Flux model GGUF Schnell: https://huggingface.co/city96/FLUX.1-...

t5-v1_1-xxl-encoder-gguf = https://huggingface.co/city96/t5-v1_1...

clip_l.safetensors = https://huggingface.co/comfyanonymous...

VAE: https://huggingface.co/black-forest-l...

ComfyUI GGUF plugin: https://github.com/city96/ComfyUI-GGUF


00:00 Intro
00:18 Method 1 Getting files
01:55 Method 1 moving files
03:00 Method 1 configuring ComfyUI
03:42 Method 2 Getting model
04:21 Method 2 moving file
04:47 Method 2 configuring ComfyUI
06:43 Method 3 Getting files
07:54 Method 3 moving files
08:49 Method 3 Installing GGUF module
10:26 Method 3 configuring ComfyUI
13:38 Ending