Model Here:  https://civitai.com/models/141592/pixelwave

Step 0: Install the custom node  GGUF loader

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

Step 1: Download two CLIP "LLMs"

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

Specifically t5xxl_fp8_e4m3fn.safetensors and  clip_l.safetensors

Clip_G = Global for the prompt you want to focus on everything

Clip_L = Local for details to specific areas of the picture lets say eyes, or nose or something and then the normal prompt So you can use 3 prompts that feed into your KSampler You type your regular prompt in clip_g you can put detailing info you want applied to all of the picture in clip_l you put specific parts of the image you want to detail further.

Clip_L : it's the industry standard if you will almost all other Clip stuff was derived from that.

Additional t5 clip encoders

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

L and G encoders

https://huggingface.co/stabilityai/stable-diffusion-3-medium/tree/main/text_encoders https://huggingface.co/stabilityai/stable-diffusion-3.5-medium/tree/main/text_encoders

Additional doc

https://comfyanonymous.github.io/ComfyUI_examples/flux/