Tencent has released Hunyuan Video, a new text-to-video generation tool powered by its Hunyuan Foundation model. The tool is designed for both enterprise and individual users.
Hunyuan Video can generate videos up to five seconds long with a resolution of 1280x720 (720p). Tencent claims its advanced 3D VAE architecture ensures visual consistency and smooth, natural motion.
Using a Multimodal Language Model (MLLM) based text encoder, Hunyuan Video reportedly outperforms existing models with text alignment rates of up to 68.5% and visual quality scores of 96.4%. Text alignment measures how accurately the generated video matches the user's text prompt.
The tool is open-source and operates on a credit system. Generating one video costs 15 credits. Credit bundles are available for purchase:
- 150 credits: $9.99 (€9.48/£7.85)
- 750 credits: $44.99 (€42.70/£35.34)
- 1500 credits: $79.99 (€75.93/£62.83)
AI video generation is a rapidly growing field, with major tech companies like OpenAI, Google, Meta, Amazon, Kuaishou, and Alibaba actively involved.
