I have made it on big hurry and sleepy just because I've being asked, so don't be too picky.

Core model for Flexible Photo Recrafting While Preserving Your Identity is of Bytedance Inc., you can read on their GitHub page all related and the License terms.

The Node for ComfyUI is of ZenAI Team, please, follow all their instructions for downloading and installing i just replaced the model and TE nodes for GGUF versions, added LoRA node and Save image node (v. 4.0 - added Switch between SIMILARITY and FLEXIBILITY models, v. 5.0 - Lora node replaced for Multiple Lora node) and made other minor adjustments.

The workflow was tested on 16Gb RTX 4080 Super with the medium speed of 2 minutes per 1024x1024 image generation.

I am very thankful to Bytedance Inc y ZenAI Team for their support and their great work, when I have a bit more time I will adjust this page accordingly.

You can use this workflow with any of my LoRAs listed in Suggested Resources

This workflow is completely free of charge, but if you want to support me for any reason and motivate me to continue my work you can buy me a coffee https://ko-fi.com/cocosus

If you get any error on python level the best way to solve it is reading history of issues on developers pages (links above) and if it persists opening a new issue. In a specific case of getting "facexlib" or "filterpy" error read please this issue and this issue how to solve it or write me, i will be glad to help in what is my own experience with this node and model, but i am not a developer of this node/model, just a workflow optimizer in this case.