Hetzner Experiments Inference API — Qwen3.8-27B — Event History

New canonical model-route record. Official Hetzner Experiments docs advertise Qwen3.8-27B with 262,144 context tokens and $0 input/output token charges while experimental. Limits: 10 requests/minute, 4M input and 100k output tokens/minute per API key; no SLA or guaranteed availability. Account-level entitlement and live authenticated /v1/models not tested. No credit balance is issued.