hunyuan3d-2.1-mlx

dgrauet/hunyuan3d-2.1-mlx

MLX format conversion of tencent/Hunyuan3D-2.1.

Converted with mlx-forge.

License

These weights are a derivative work and are distributed under the same terms as the model they were converted from. A verbatim copy ships with them as LICENSE and Notice.txt, and the upstream text is at https://github.com/Tencent-Hunyuan/Hunyuan3D-2.1/blob/main/LICENSE. Use of these weights is subject to it, including any restrictions it attaches.

Usage

These weights can be used with Hunyuan3D-2.1-mlx:

pip install mlx mlx-arsenal safetensors Pillow trimesh scikit-image PyMCubes scipy
git clone https://github.com/dgrauet/Hunyuan3D-2.1-mlx.git && cd Hunyuan3D-2.1-mlx

python - <<'PY'
from hy3dshape.hy3dshape.pipeline_mlx import ShapePipeline

pipe = ShapePipeline.from_pretrained("dgrauet/hunyuan3d-2.1-mlx")
mesh = pipe("your_image.png", num_inference_steps=50, guidance_scale=7.5, octree_resolution=256)
mesh.export("output.glb")
PY

Related Projects

• Code: https://github.com/dgrauet/Hunyuan3D-2.1-mlx
• Upstream model: https://huggingface.co/tencent/Hunyuan3D-2.1
• Paper: https://arxiv.org/pdf/2506.15442

Files

• LICENSE (17.50 KB)
• Notice.txt (16.25 KB)
• config.json (1.29 KB)
• dit.safetensors (5.68 GB)
• image_encoder.safetensors (580.59 MB)
• paint_clip.safetensors (1.18 GB)
• paint_dino.safetensors (2.12 GB)
• paint_unet.safetensors (3.66 GB)
• paint_vae.safetensors (159.58 MB)
• realesrgan_x4plus.safetensors (63.76 MB)
• split_model.json (1.15 KB)
• vae.safetensors (625.17 MB)