hunyuan3d

name	hunyuan3d
description	Generate 3D models from images using Tencent's Hunyuan3D-2.1 via Hugging Face Gradio API. Supports single-image and multi-view input, geometry-only or textured output, and GLB/OBJ/PLY/STL export. USE WHEN converting a 2D image or photo to a 3D mesh, or generating textured 3D assets.
cluster	media-gen
version	1.0.0

Hunyuan3D — Image-to-3D Model Generation

Generate 3D models from images using Tencent's Hunyuan3D-2.1 hosted on Hugging Face Spaces. Supports geometry-only and fully textured output with export to multiple 3D formats.

Prerequisites

Ensure gradio_client is installed:

pip install gradio_client

Workflow

Quick Generation (Single Command)

To generate a 3D model from an image, run the bundled script:

# Geometry + Texture (default)
python scripts/hunyuan3d_generate.py --image <path-to-image> --mode full --output ./output/

# Geometry only (faster)
python scripts/hunyuan3d_generate.py --image <path-to-image> --mode shape --output ./output/

Multi-View Generation

When multiple views of the subject are available, provide them for higher quality:

python scripts/hunyuan3d_generate.py \
    --front front.png --back back.png --left left.png --right right.png \
    --mode full --output ./output/