Suite of models to evaluate the image quality of text-to-image models with respect to their input prompts.