Flux 2 is a state of the art image diffusion model.
Text encoder file: mistral_3_small_flux2_fp8.safetensors (goes in ComfyUI/models/text_encoders/).
Fp8 diffusion model file: flux2_dev_fp8mixed.safetensors (goes in ComfyUI/models/diffusion_models/). If you want the full sized diffusion model you can find the flux2-dev.safetensors on the official repo here
VAE: flux2-vae.safetensors (goes in ComfyUI/models/vae/)
You can then load up or drag the following image in ComfyUI to get the workflow:
This model supports multiple reference images as optional inputs. This example workflow has two of them bypassed out but you can add more.