HunyuanVideo is a new open-source AI model for video creation that Tencent released to match the capabilities of current commercial solutions. Tencent claims it is the largest publicly available model, with over 13 billion parameters. Technical material claims that HunyuanVideo outperforms three key Chinese video-generating models and existing systems, such as Runway Gen-3 and Luma 1.6. In tests of motion quality, the system performs exceptionally well.