awesome-free-ai-coding · every free model · every provider · filter the list

Hetzner Inference API free tier

🔌 LLM APIs with free tier · no card · live — last verified by a probe on 2026-11-01 · docs.hetzner.com · back to the whole list

What you get

OpenAI-compatible API on Hetzner’s own EU hardware, free for as long as the experiment runs

Free models

qwen3.6, qwen3.8-27b

Limits, in the vendor’s words

4,000,000 input tokens/60 seconds; 100,000 output tokens/60 seconds; 10 requests/60 seconds per key

Free while the Inference API remains experimental; Hetzner says it will email advance notice of a change. Per-key request and token windows both apply, with HTTP 429 at a cap. No daily, monthly or lifetime cap or end date is published. The service is offered as is for experimental use, without guaranteed availability or backups. Minting a token needs a Hetzner account; payment verification can include a card charge or other routes.

Where it is offered

The vendor names no country it keeps the offer from (source, read 2026-09-26).

What happens to what you send

What you send is not used to train models. In the vendor’s words: “Prompts and outputs: Under our zero-retention policy, we do not store them. They are also not used in any other way to train or improve the AI models.” (source).

Connect

• Base URL: https://inference.hetzner.com/api/v1
• Key: HETZNER_INFERENCE_API_KEY — get one at https://experiments.hetzner.com/inference
• Callable ids: Qwen/Qwen3.6-35B-A3B-FP8, Qwen3.8-27B
• Note: both ids are 262K-context and take text and images; /v1/models, /v1/completions and /v1/chat/completions are the whole surface. The key is minted in the Inference tab of experiments.hetzner.com, which is a client-rendered page behind a Hetzner login — the readable copy of the terms is the docs page this row probes

Try it from your terminal with your key in HETZNER_INFERENCE_API_KEY — it goes from your machine to the vendor and nowhere else:

curl -s https://inference.hetzner.com/api/v1/chat/completions \
  -H "Authorization: Bearer $HETZNER_INFERENCE_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"model":"Qwen/Qwen3.6-35B-A3B-FP8","messages":[{"role":"user","content":"2+2? MAKE NO MISTAKES."}]}'

Evidence

• Probe: the page at https://docs.hetzner.com/general/company-and-policy/experiments/inference/, anchored on remains in experimental status, it is free of charge, https://inference.hetzner.com/api/v1
• Source: https://docs.hetzner.com/general/company-and-policy/experiments/inference/
• Source: https://experiments.hetzner.com/inference

History

Each line is a change to what this page publishes, dated the day it reached the list, in UTC.

• 2026-09-17 — Free models changed: added qwen3.8-27b; dropped qwen3.8
• 2026-08-30 — Added: OpenAI-compatible API on Hetzner’s own EU hardware, free for as long as the experiment runs

---

Generated from registry.yaml on 2026-11-02 and re-verified twice a week; the full list, the Atom feed and the machinery are at https://github.com/mvalentsev/awesome-free-ai-coding.

Something wrong or missing? Suggest a service or a change — or read how the list works.