Tencent’s Hunyuan team unveiled a cutting-edge artificial intelligence (AI) model on Sunday, capable of generating immersive three-dimensional (3D) environments. Named Hunyuan 3D World Model 1.0, this new offering is open-source and accommodates both textual and visual inputs. Users can create 360-degree worlds that are interactive and explorable, with the capability to navigate through these environments as playable assets. The tech giant emphasizes that this model can significantly enhance video game development, virtual reality experiences, and digital content production.
Hunyuan 3D World Model 1.0 Developed on Innovative Framework
The realm of world generation models represents a specialized segment of generative AI, focusing on producing geometrically consistent environments with efficient rendering and visual diversity. The environments crafted through these models find utility in various applications, from video game design to virtual robot training.
The Hunyuan 3D World Model is accessible for download via Tencent’s GitHub and Hugging Face listings. It is released under a permissive “tencent-hunyuanworld-1.0-community” license, permitting both academic research and commercial usage. Users interested in testing the model without local installation can visit Tencent’s platform to start generating worlds.
We’re thrilled to release & open-source Hunyuan3D World Model 1.0! This model enables you to generate immersive, explorable, and interactive 3D worlds from just a sentence or an image.
It’s the industry’s first open-source 3D world generation model, compatible with CG pipelines… pic.twitter.com/CpETdVO7vW
— Hunyuan (@TencentHunyuan) July 27, 2025
According to information on the project page, traditional world generation techniques have primarily concentrated on either video-centric or 3D-centric methodologies. While video-based methods excel in visual variety, 3D methods deliver consistency and rendering efficiency. The Hunyuan team asserts that neither approach offers the comprehensive benefits of both, leading to the creation of their innovative framework.
This new framework employs a semantically layered 3D mesh representation to project panoramic images as 360-degree world proxies. These proxies are then dissected and reconstructed, enabling the creation of diverse worlds. Utilizing this framework, researchers have trained the model to generate coherent and interactive 3D environments.
The Hunyuan team noted that the open-source version of this AI model is influenced by Flux’s image generator, yet can be adapted for other models, including Hunyuan Image, Kontext, Stable Diffusion, and others. However, specific information regarding the availability of the model through application programming interfaces (APIs) has not been disclosed.
