On December 17, 2025, Tencent Hunyuan Lab proudly announced the launch of its latest innovation—the HunyuanVideo 1.5 video generation model. This cutting-edge model is built upon the Diffusion Transformer architecture, boasting an impressive 8.3 billion parameters. It is capable of generating high-definition videos ranging from 5 to 10 seconds in length. What's more, it can operate seamlessly on consumer-grade graphics cards equipped with just 14GB of memory, making it accessible to a wider range of users.

One of the standout features of HunyuanVideo 1.5 is its versatility in input methods. Users have the flexibility to generate videos by providing either text descriptions or image prompts. Additionally, the model supports both Chinese and English languages, catering to a global user base. It also accommodates multiple styles and complex instructions, allowing for a high degree of customization and creativity in video generation.

The model has already been integrated into the "Yuanbao" platform, providing users with a user-friendly interface to explore its capabilities. Furthermore, in a move that underscores Tencent's commitment to open innovation, the model weights have been made available as open-source, enabling researchers and developers worldwide to contribute to its ongoing development and refinement.