Run FLUX.1 with ComfyUI on M3 Mac

What is FLUX.1

Announcing Black Forest Labs

Install ComfyUI

git clone https://github.com/comfyanonymous/ComfyUI.git
cd ComfyUI

# create venv
python -m venv venv

# activate venv
source venv/bi/activate

# install packages
pip install torch==2.3.1 torchaudio==2.3.1 torchvision==0.18.1
pip install -r requirements.txt

Run ComfyUI

python main.py --listen

If everything works well, you will see the following

Starting server

To see the GUI go to: http://0.0.0.0:8188

Download Flux, CLIP, VAE.

Flux.1

black-forest-labs/FLUX.1-dev at main

After downloading flux1-dev.sft, you need to move the file into ComfyUI/models/unet

CLIP

comfyanonymous/flux_text_encoders at main

Download t5xxl_fp16.safetensors, clip_l.safetensors, and t5xxl_fp8_e4m3fn.safetensors. Then move them into ComfyUI/models/clip

VAE

ae.sft · black-forest-labs/FLUX.1-schnell at main

Download ae.sft and move it into ComfyUI/models/vae

Rerun ComfyUI and Import workflow

python main.py --listen

Flux under 12GB VRAM | Civitai

Download the image and click Load button and load the image you download. Then you will see the workflows to generate an image with Flux.1

If your Mac memory is less than 32GB, you may not be able to generate a high-quality image. Actually I couldn’t generate a nice quality image 😭

Useful resources

ComfyUI workflow for Flux (simple) | ComfyUI Workflow | OpenArt

Flux Examples