Install ComfyUI or update to latest version

Download ae.sft from https://huggingface.co/black-forest-labs/FLUX.1-dev/tree/main and move it to .../ComfyUI/models/vae/

Download flux1-dev.sft from https://huggingface.co/black-forest-labs/FLUX.1-dev/tree/main and move it to .../ComfyUI/models/unet/

Download clip_l.safetensors from https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main and move it to ../ComfyUI/models/clip/

Download t5xxl_fp8_e4m3fn.safetensors from https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main and move it to .../ComfyUI/models/clip/

Download flux_dev_example.png from https://github.com/comfyanonymous/ComfyUI_examples/tree/master/flux

add "--lowvram" to your startup parameters

find the "Load Diffusion Model" node (upper left corner) and set "weight type" to "fp8-e4m3fn"

find the "DualClipLoader"-node (upper left corner) and set "clip_name1" to "t5xxl_fp8_e4m3fn.safetensors"

click "queue prompt" (or change the prompt before in the "CLIP Text Encode (Prompt)"-node

---------------------------

Discord:  https://discordapp.com/users/339415040308412418

Any support is appreciated, thank you very much :)

patreon.com/akihungac

buymeacoffee.com/akihungac

paypalme/manhtuanphoto