Create original images from text prompts, sketches, or reference images.
No models in this category are strong on that dimension yet.