MindPatternSubscribe

Source trail

Hetzner Experiments (analysis: sliplane.io)

Public MindPattern findings, entities, and graph evidence that cite this source.

Findings

1

All-time hits

1

High value

0

Last seen

2026-07-24

Related findings

2026-07-24 / REDDITHetzner Is Quietly Testing Free OpenAI-Compatible Inference, Starting With Qwen3.6-35BSurfacing on Hacker News July 24 (111 points), Hetzner has an OpenAI-compatible inference API running on its own infrastructure via the Hetzner Experiments dashboard — free, token-authenticated, with explicitly no billing, no SLA and no production guarantee. Only one model is live: Qwen3.6-35B, a MoE with quantized weights, measured informally at ~153ms median time-to-first-token and 224 output tokens/sec. The open question is hardware: Hetzner's public GPU lineup is RTX 4000 Ada and RTX PRO 6000 Blackwell, workstation-class cards that won't serve frontier-size models — whether they buy datacenter GPUs is the real signal about a European budget inference tier.

Open latest cited source