Deploy PyTorch models
to the edge

Quantize, fuse, and optimize your models for edge inference. Visualize every layer before you ship.

Open Visualizer View Docs

pip install embedl-deploy

Model Gallery

Explore optimized models — click any card to inspect the full computation graph

25.6M params · 55 fusions

ViT-B-16 Transformer

86.6M params · 87 fusions

ConvNeXt-Tiny CNN

28.6M params · 78 fusions

Swin-T Transformer

28.3M params · 95 fusions