Zinc

What it does

Zig INferenCe Engine — Local LLM inference on AMD GPUs and Apple Silicon

In their own words: Local LLM inference for consumer GPUs: AMD Radeon, Intel Arc, Apple Silicon, and NVIDIA RTX CUDA experiments. The current strongest measured path is AMD RDNA4, while Intel Arc is now an official Linux Vulkan support target.

It's a more tools tool from Zolotukhin. A free tier lets you try it before paying — useful when you're keeping costs low. Visit the website for the latest features and pricing.