Tencent has reportedly unveiled a suite of five new artificial intelligence (AI) models as the tech giant focuses on improving its AI-supported 3D capabilities.
Based on Hunyuan3D-2.0 technology, the company’s advanced AI system for generating high-resolution textured 3D assets from text and images, the new tools can convert text and images into 3D, Reuters reported on Tuesday.
The new models, which are designed for game designers and developers, are all built on open-source systems with an aim of enabling community collaboration and transparency.
They include a new “turbo” version capable of generating 3D images in 30 seconds while maintaining high image quality.
The new “turbo” model builds on the Hunyuan Turbo S, a model that Tencent released in February that was designed to offer higher speed than previous models, such as DeepSeek R1, with a significant reduction in latency, while equalling the performance of advanced models such as DeepSeek V3.
Launched in January, the Hunyuan3D-2.0 technology aims to solve the complexities of shape and texture generation and produce photorealistic 3D models with detailed geometry and vivid textures.
Tencent first unveiled its 3D AI models in November 2024 and has continued to work on developments in the field of AI together with other Chinese companies, including Huwaei and DeepSeek, as they seek to win competition in the US market by offering higher performance models at a lower cost.
