Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AI

How-To How to actually use this

What changed: Hugging Face released @huggingface/kernels, a package offering 200+ WebGPU kernels to run AI models in browsers.

How to use it:

1. Install the @huggingface/kernels package in your web project.
2. Use the kernels to run AI models directly in the browser via WebGPU.
3. Test on a WebGPU-compatible browser and device for local inference.

Good for: developers running AI models locally in browsers via WebGPU.