Tencent launches Hy4 with visual understanding and computer-use features

Tencent launched Hy4 on November 4, adding visual understanding and computer-use capabilities to its flagship AI model, according to TechNode (China Tech). The move underscores Tencent's bid to pair multimodal reasoning with hands-on software control to streamline complex digital work. It lands as Chinese platforms race to push AI deeper into productivity and design tools amid rising competition and policy scrutiny at home and abroad.

TechNode reports Hy4 can interpret charts and documents, reason across multiple images, and operate both command-line tools and graphical interfaces. Targeted training includes applications such as Blender and CAD software, and tasks that span multiple programs and workflows, aiming at creative and engineering use cases. Watch for announcements on availability, public developer access, and benchmarks against rival multimodal agents. [Developing. This article will be updated.]