Omnigen 2 is a model that can be used to edit images with text prompts.
You will first need:
omnigen2_fp16.safetensors goes in: ComfyUI/models/diffusion_models/
qwen_2.5_vl_fp16.safetensors goes in: ComfyUI/models/text_encoders/
ae.safetensors, this is the flux VAE that you might already have, it goes in: ComfyUI/models/vae/
This is a basic workflow using an image as a character reference. For multiple image inputs chain ReferenceLatent nodes together
You can load this image in ComfyUI to get the full workflow.
You can find the input image here
