Hetzner is working on LLM Inference

A first look at Hetzner's experimental LLM inference API, which is fast and free to use, but its long-term viability depends on the company's ability to provide high-performance hardware.