Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Hunyuan3D 2.5: Towards High-Fidelity 3D Assets Generation with Ultimate Details

Zeqiang Lai, Yunfei Zhao, Haolin Liu, Zibo Zhao, Qingxiang Lin, Huiwen Shi, Xianghui Yang, Mingxin Yang, Shuhui Yang, Yifei Feng, Sheng Zhang, Xin Huang, Di Luo, Fan Yang, Fang Yang, Lifu Wang, Sicong Liu, Yixuan Tang, Yulin Cai, Zebin He, Tian Liu, Yuhong Liu, Jie Jiang, Linus, Jingwei Huang, Chunchao Guo

TL;DR

Hunyuan3D 2.5 addresses the challenge of high-fidelity 3D asset generation by introducing the LATTICE shape foundation model and a BRDF-consistent, multi-view texture framework that includes learnable material embeddings and a dual-phase resolution strategy. The two-stage pipeline yields detailed geometry and realistic PBR textures, achieving state-of-the-art results in both shape fidelity and end-to-end texture realism, validated across quantitative metrics and user studies. This work significantly advances practical 3D content creation for industries such as games, film, and AR/VR by delivering smooth geometry with accurate texture maps and realistic lighting behavior. The approach balances scalability, efficiency, and quality, enabling high-resolution, asset-wide consistency in complex scenes.

Abstract

In this report, we present Hunyuan3D 2.5, a robust suite of 3D diffusion models aimed at generating high-fidelity and detailed textured 3D assets. Hunyuan3D 2.5 follows two-stages pipeline of its previous version Hunyuan3D 2.0, while demonstrating substantial advancements in both shape and texture generation. In terms of shape generation, we introduce a new shape foundation model -- LATTICE, which is trained with scaled high-quality datasets, model-size, and compute. Our largest model reaches 10B parameters and generates sharp and detailed 3D shape with precise image-3D following while keeping mesh surface clean and smooth, significantly closing the gap between generated and handcrafted 3D shapes. In terms of texture generation, it is upgraded with phyiscal-based rendering (PBR) via a novel multi-view architecture extended from Hunyuan3D 2.0 Paint model. Our extensive evaluation shows that Hunyuan3D 2.5 significantly outperforms previous methods in both shape and end-to-end texture generation.

Hunyuan3D 2.5: Towards High-Fidelity 3D Assets Generation with Ultimate Details

TL;DR

Hunyuan3D 2.5 addresses the challenge of high-fidelity 3D asset generation by introducing the LATTICE shape foundation model and a BRDF-consistent, multi-view texture framework that includes learnable material embeddings and a dual-phase resolution strategy. The two-stage pipeline yields detailed geometry and realistic PBR textures, achieving state-of-the-art results in both shape fidelity and end-to-end texture realism, validated across quantitative metrics and user studies. This work significantly advances practical 3D content creation for industries such as games, film, and AR/VR by delivering smooth geometry with accurate texture maps and realistic lighting behavior. The approach balances scalability, efficiency, and quality, enabling high-resolution, asset-wide consistency in complex scenes.

Abstract

In this report, we present Hunyuan3D 2.5, a robust suite of 3D diffusion models aimed at generating high-fidelity and detailed textured 3D assets. Hunyuan3D 2.5 follows two-stages pipeline of its previous version Hunyuan3D 2.0, while demonstrating substantial advancements in both shape and texture generation. In terms of shape generation, we introduce a new shape foundation model -- LATTICE, which is trained with scaled high-quality datasets, model-size, and compute. Our largest model reaches 10B parameters and generates sharp and detailed 3D shape with precise image-3D following while keeping mesh surface clean and smooth, significantly closing the gap between generated and handcrafted 3D shapes. In terms of texture generation, it is upgraded with phyiscal-based rendering (PBR) via a novel multi-view architecture extended from Hunyuan3D 2.0 Paint model. Our extensive evaluation shows that Hunyuan3D 2.5 significantly outperforms previous methods in both shape and end-to-end texture generation.

Paper Structure

This paper contains 14 sections, 1 equation, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Detailed Shape Generation
  4. Realistic Texture Generation
  5. Multi-Channel Material Generation.
  6. Geometric Alignment.
  7. Evaluation
  8. Shape Generation
  9. Texture Generation
  10. Related Work
  11. Shape Generation.
  12. Texture Synthesis.
  13. Conclusion
  14. Contributors

Figures (8)

  • Figure 1: High quality 3D assets generated by Hunyuan3D 2.5.
  • Figure 2: Drawbacks of existing methods: failure at detail generation and incorrect PBR.
  • Figure 3: Overview of Hunyuan3D 2.5 pipeline. It separates the 3D asset generation into two stages: first, it generates the shape, and then it creates the texture based on that shape.
  • Figure 4: Illustration of major features of the new shape generation model in Hunyuan3D 2.5.
  • Figure 5: Overview of material generation framework.
  • ...and 3 more figures