This workflow does the following:<ul><li>Load 1-100 random images from a selected directory</li><li>Create a grid from the loaded images</li><li>Using <a target="_blank" rel="ugc" href="https://github.com/Creepybits/ComfyUI-Creepy_nodes">Gemini</a> to create a prompt by using the grid as a single image</li><li>Use an input image as style reference</li><li>Creates 4 different images using <a target="_blank" rel="ugc" href="https://github.com/Jonseed/ComfyUI-Detail-Daemon">Detail Daemon</a></li><li>Using <a target="_blank" rel="ugc" href="https://github.com/Zuellni/ComfyUI-PickScore-Nodes">Pick Score</a> to select the image that best reflects the prompt</li><li>Upscale image using an upscale model</li><li>Refines the image using SDXL</li><li>Go through Face Detailer</li><li>Saves the image to <a target="_blank" rel="ugc" href="https://www.patreon.com/posts/129034256">Google Drive</a> or <a target="_blank" rel="ugc" href="https://www.patreon.com/posts/129775871">OneDrive</a></li></ul>

Files with SHA256: fc24cdb5d0db6e2df2ae860d37165da6811a524529139b8d93f971c40d390442