Random historical mashup to test how diferent SDXL models deals with multiple face generation of multiple subjects. WyvernMix really won using the SeargeSDXL full workflow with Jugggernaut Cinematic Lora.
