Using a reference image (person, object) with Krea2 is something that needs a bit more than just adding a reference image and that's it. So, a simple workflow, that uses 1 reference image and it's called to several separate prompts (or one). Each plate from Multy_prompts (custom node <a target="_blank" rel="ugc" href="https://github.com/JDJoe/ComfyUI-Multy_prompts">https://github.com/JDJoe/ComfyUI-Multy_prompts</a>) is an edit of the same reference photo. The photo holds the person. The plate says what to add. Latest version can be found here: <a target="_blank" rel="ugc" href="https://github.com/JDJoe/krea2-ref-image-multy-prompts">https://github.com/JDJoe/krea2-ref-image-multy-prompts</a> Download the .json and you are good to go. Make sure you custom nodes installed! NB! You must have krea2_identity_edit_v1_2.safetensors or later version. Use what ever other LoRA's you like.

Files with SHA256: 695532c247ca461e4cdf0d8f762b56211c14f0040b8f115702706aff3888ad45