Alibaba has recently introduced Tongyi Wanxiang Wan2.2, a pioneering open-source video generation model designed to produce film-quality content. This model seamlessly integrates three essential cinematic aesthetic components: lighting and shadow, color, and camera language. By supporting over 60 adjustable parameters, Wan2.2 dramatically accelerates the process of creating high-quality videos. Currently capable of generating 5-second HD videos at a time, the model holds promise for future enhancements to increase video duration. Wan2.2 has made available three open-source models: text-to-video, image-to-video, and unified video generation. All three models are industry firsts to employ the Mixture of Experts (MoE) architecture, resulting in a 50% reduction in computing resources. Users can effortlessly create diverse film-grade scenes by simply inputting combinations of keywords.