Same thing with image model. 4 Go stable diffusion model can draw and represent anything humanity know.