MagicAnimate is an advanced framework based on diffusion models for human image animation. It can generate animation videos from a single image and dynamic videos, maintaining temporal consistency, preserving the features of the reference image, and significantly enhancing the animation's fidelity. MagicAnimate supports image animation using motion sequences from various sources, including cross-identity animations and unseen domains such as oil paintings and film characters. It also integrates seamlessly with T2I diffusion models like DALLE3, enabling dynamic actions to be applied to text-generated images. MagicAnimate was jointly developed by Show Lab at National University of Singapore and ByteDance.