A marketing team wants to create original promotional visuals by typing descriptive phrases such as 'a red sports car driving through a mountain valley at sunset.' They have no existing images to work from and want the system to produce brand-new pictures. Which type of AI model should they use?