Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets Generation

2025-01-21

AI Summary

Existing 3D generative methods suffer from significant bottlenecks in geometric detail fidelity and texture realism. To address this, we introduce Hunyuan3D—the first end-to-end high-resolution textured 3D asset generation framework. It comprises two core components: (1) Hunyuan3D-DiT, a geometry generation model based on a scalable streaming diffusion Transformer for precise multimodal conditional alignment; and (2) Hunyuan3D-Paint, a texture synthesis model integrating geometric priors with 3D-aware diffusion priors to enable high-fidelity UV mapping at resolutions ≥1024×1024. Complementing these, Hunyuan3D-Studio is an open-source platform supporting interactive editing and animation. Extensive evaluation demonstrates state-of-the-art performance across geometric accuracy, conditional consistency, and texture photorealism. We fully open-source the models, training code, and pre-trained weights, establishing the first large-scale foundational 3D generative model ecosystem.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

We present Hunyuan3D 2.0, an advanced large-scale 3D synthesis system for generating high-resolution textured 3D assets. This system includes two foundation components: a large-scale shape generation model -- Hunyuan3D-DiT, and a large-scale texture synthesis model -- Hunyuan3D-Paint. The shape generative model, built on a scalable flow-based diffusion transformer, aims to create geometry that properly aligns with a given condition image, laying a solid foundation for downstream applications. The texture synthesis model, benefiting from strong geometric and diffusion priors, produces high-resolution and vibrant texture maps for either generated or hand-crafted meshes. Furthermore, we build Hunyuan3D-Studio -- a versatile, user-friendly production platform that simplifies the re-creation process of 3D assets. It allows both professional and amateur users to manipulate or even animate their meshes efficiently. We systematically evaluate our models, showing that Hunyuan3D 2.0 outperforms previous state-of-the-art models, including the open-source models and closed-source models in geometry details, condition alignment, texture quality, and etc. Hunyuan3D 2.0 is publicly released in order to fill the gaps in the open-source 3D community for large-scale foundation generative models. The code and pre-trained weights of our models are available at: https://github.com/Tencent/Hunyuan3D-2

Problem

• 3D model generation
• detail processing
• texture handling

Innovation

• High-Detail 3D Modeling
• Texture Enhancement
• User-Friendly Platform