TMTPOST — Kunlun Wanwei, a leading Chinese Internet company, has unveiled China’s first open-source AI model designed for short-drama video creation, SkyReels-V1. Along with it, the company introduced SkyReels-A1, the country's first SOTA-level emotion and action-controllable algorithm based on video base models. SkyReels-V1 is capable of generating cinematic-level micro-expressions and character performances. It supports 33 detailed facial expressions and over 400 natural actions, enabling realistic emotional expressions that closely mimic human performance. Not only does SkyReels-V1 support text-to-video generation, but it can also generate videos from images. With its vast parameters, it stands out as the largest open-source video generation model capable of generating video from images. At the same resolution, SkyReels-V1 achieves open-source SOTA (state-of-the-art) performance across multiple metrics. This breakthrough positions Kunlun Wanwei at the forefront of AI-driven video content creation in China, offering a powerful tool for the entertainment industry and beyond.