Instructions to use ddalcu/Pixal3D-MLX-Serve-8bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use ddalcu/Pixal3D-MLX-Serve-8bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download ddalcu/Pixal3D-MLX-Serve-8bit --local-dir Pixal3D-MLX-Serve-8bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Pixal3D for mlx-serve (8-bit)
TencentARC/Pixal3D repackaged for mlx-serve
on Apple Silicon. Tencent's image-to-3D model, built on the TRELLIS.2 backbone with pixel-aligned conditioning: an image with a transparent background becomes a textured PBR GLB. Single-view and multi-view (*_mv) models.
- The flow-model (DiT) block projections are affine-quantized to 8 bit, group size 64; everything else is as released.
- Includes the DINOv3 image encoder it conditions on (float16); its licence is in
dinov3/LICENSE.md. - No background remover is included: send an image with a transparent background (alpha is the subject).
Licences
The Pixal3D weights are MIT (LICENSE). The bundled DINOv3 encoder is Meta's, under the DINOv3 License
(dinov3/LICENSE.md), which travels with any redistribution of it.
The NAF feature upsampler in naf/ is Apache-2.0 (naf/LICENSE).
Run: mlx-serve --model-dir <folder holding this repo>, then POST /v1/3d/generations.
Converted by tests/convert_trellis2_pixal3d.py in the mlx-serve repository.
Hardware compatibility
Log In to add your hardware
Quantized
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support
Model tree for ddalcu/Pixal3D-MLX-Serve-8bit
Base model
TencentARC/Pixal3D