The Chinese giant said it is open sourcing four models that are part of its Wan2.1 series, the latest version of the company's foundational AI model that can generate images and video from text and ...