Hetzner experiments with open-weight LLM inference
Hetzner testing a free, no-SLA open-weight inference API is a useful signal that commodity infra providers are probing hosted LLM inference as an infrastructure product, not only renting GPUs.
Links: Original source · Shared link · Related link · Related link 1 · Related link 2
Logged at IST: 2026-08-11 20:03 IST
What it is: Hetzner's exploratory Experiments platform with an experimental open-weight LLM inference API.
Gist: Hetzner says the API is free for now, has no SLAs, and is meant for users to test their own workloads and report what works. They are explicit that this may not become a permanent product.
The linked site resolves to experiments.hetzner.com; the app exposes an AI Inference · Hetzner Experiments area, API-token creation, docs links, and an inference endpoint at https://inference.hetzner.com/api/v1. The unauthenticated /models endpoint returns 401 unauthorized, so the actual model list was not visible without a token.
Newsletter angle: Good ai-infra item: a traditional hosting provider is testing LLM inference as a managed API, not just selling GPU servers. The cautious free, no SLA, maybe not permanent framing makes it useful market signal rather than a finished product launch.