Advanced AI Art Remix Workflow for FLUX

Use this ComfyUI workflow to remix and blend art styles, control depth and lighting, and emulate artist techniques using multiple reference images and style choices; it generates AI artwork and can upscale it to high-resolution outputs.

In ComfyUI, load the workflow, add reference images for style, depth, and lighting, choose additional styles, generate artwork, then upscale the result.

Inputs include multiple reference images for style, depth, and lighting, plus additional style selections.

The workflow generates AI artwork and supports high-resolution upscaled outputs.

Install the listed model files and node packs before loading the workflow in ComfyUI.

This is a ComfyUI workflow; no Pipe version is provided.

The model files are 4xUltrasharp_4xUltrasharpV10.pth, clip_l.safetensors, flux1-dev.safetensors, t5xxl_fp16.safetensors, and vae.safetensors.

Node packs include ComfyUI_Comfyroll_CustomNodes, comfyui-custom-scripts, comfyui_layerstyle, and comfyui_ultimatesdupscale.

Other node packs are comfyui-ollamagemini, rgthree-comfy, and wavespeed.

Choose additional styles with the All-IN-ONE-style node.

The workflow integrates ComfyUI-OllamaGemini for AI-powered text generation and analysis.

It uses Flux1-dev and Flux1-schnell models for base generation.

Before setup, verify that vae.safetensors is the VAE file you intend to use and that 4xUltrasharp_4xUltrasharpV10.pth is the upscaler file you intend to use.

Suggestion · not verified

Before loading the workflow, verify how the wavespeed node pack should be installed.

Suggestion · not verified