HunyuanVideo is a systematic framework for training large video generation models, open-sourced by Tencent. By leveraging key technologies such as data curation, joint image-video model training, and efficient infrastructure, it successfully trained a video generation model with over 13 billion parameters, making it the largest among all open-source models. HunyuanVideo excels in visual quality, motion diversity, text-video alignment, and generation stability, surpassing several industry-leading models including Runway Gen-3 and Luma 1.6. With open-source code and model weights, HunyuanVideo aims to bridge the gap between closed-source and open-source video generation models, promoting the vibrant development of the video generation ecosystem.