Tencent’s AI model generates virtual worlds from a single text prompt

China – Tencent has launched an open-source AI model, Hunyuan3D World Model 1.0, that allows users to generate complex 3D environments from a single text prompt or image.

The release of this new model marks a significant advance in generative AI, enabling creators to quickly build detailed 3D worlds. The tool transforms simple inputs into full 360-degree panoramic scenes, complete with editable object meshes that can be exported directly to major platforms like Unity and Unreal Engine. Tencent touts it as the first open-source model of its kind to offer such comprehensive world generation.

The Hunyuan3D suite, which also includes the asset-generation tool Hunyuan3D 2.0, now allows users not only to build entire virtual worlds but also to populate them with characters and other elements. This integration streamlines the creative process for professionals in VR, game design and digital art, offering a powerful, accessible solution for rapid content creation.

For more insights on how early adopters are using real-time AI-powered simulation engines, read The Future Laboratory’s and Together Group’s New Codes of Luxury: Immersive Technologies for Transformation report.