The ByteDance Intelligent Creation Team has launched the latest single-image video driving technology, X-Portrait 2. This portrait animation technology generates highly expressive and realistic character animations and video clips from user-provided static portrait images and driving performance videos. It significantly reduces the complexity of existing motion capture, character animation, and content creation processes. X-Portrait 2 builds a state-of-the-art expression encoder model that implicitly encodes every subtle expression in the input and is trained on large-scale datasets. This encoder is then combined with a powerful generative diffusion model to produce smooth and expressive videos. X-Portrait 2 is capable of conveying subtle and nuanced facial expressions, including challenging expressions like pouting, tongue sticking out, cheek puffing, and frowning, ensuring high-fidelity emotional conveyance in the generated videos.