First look: CSX earnings

Hetzner is experimenting with LLM inference via an OpenAI-compatible API to evaluate user demand, system scalability, and load capacity. This is not a finished product launch; currently, there is no billing, no SLA, and no production guarantee, with only one model available. Hetzner aims to learn which features matter and how the system handles load by placing this early version in front of users to observe the results and gather data.