A medium-sized text-to-image model in the Stable Diffusion 3.5 family, offering improved quality and efficiency.