Hunyuan3D
Tencent's open-source engine turning text or images into 3D models in seconds.
About
Hunyuan3D 2.0 is Tencent's open-source 3D asset generation system, billed as an industry-first fully open one-stop 3D AI creation engine. It uses a two-stage pipeline: a flow-matching diffusion transformer (Hunyuan3D-DiT) generates mesh geometry from a text prompt or single image, then Hunyuan3D-Paint synthesizes high-resolution PBR textures (albedo, normal, metallic, roughness). Turbo variants distill the pipeline down to roughly 30-second generations.
Both stages are open-weight (Apache 2.0 plus Tencent Community License), with weights on Hugging Face and code on GitHub, plus ComfyUI and Blender addon integrations. It targets game developers and designers who need production-grade meshes fast, though like all feed-forward 3D generators it outputs static meshes — animation rigs still need a DCC tool — and best results need a capable CUDA GPU.
What it can do
- Text-to-3D mesh generation
- Image-to-3D mesh generation
- PBR texture synthesis (albedo, normal, metallic, roughness)
- Turbo variants for ~30s generation
- GLB/OBJ export with clean topology
- Blender addon and ComfyUI integration
Setup
Clone the official Tencent-Hunyuan/Hunyuan3D-2 repo and pip install requirements, or use the Hugging Face diffusers/ComfyUI integrations.
Prerequisites
CUDA GPU (weights ~15GB for shape + texture); Python 3.10+ with PyTorch.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 64; Trending score 85/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.