Awesome Video Diffusion is a curated list of diffusion models for video generation, editing, restoration, and related applications.
The repository addresses the challenge of tracking and organizing the rapidly expanding landscape of video diffusion research. It provides a structured collection of models and tools organized by capability—including video generation, editing, restoration, character and motion customization, long-form video synthesis, 3D and 4D generation, talking head synthesis, and video understanding. The list encompasses open-source toolboxes, foundation models, evaluation benchmarks, and commercial products, serving as a reference point for developers and researchers exploring what diffusion-based approaches can accomplish in video synthesis and manipulation.
Developers should use this resource when evaluating which video diffusion models or tools might suit their project needs. The list is particularly valuable for those working on video generation, controllable video synthesis, video editing, or understanding the state of the field. Rather than comparing specific models against alternatives, the repository organizes existing work by application domain, making it useful for discovering what approaches exist for a particular problem—whether that involves character customization, motion control, long-form generation, or specialized tasks like virtual try-on or talking head generation.
Maintainers typically respond to new issues and pull requests within a day.
