Free Tier · verified Aug 11, 2026
Hetzner: Free experimental OpenAI-compatible inference
Hetzner customers can create an API token for a no-cost experimental OpenAI-compatible endpoint serving Qwen3.6-35B-A3B-FP8 with text and image input. Hetzner explicitly says the service is experimental, availability is not guaranteed, and it should not be used for production.
PrototypingMultimodalOpenai Compatible
What you actually get
- Eligibility
- Hetzner customers with access to the Experiments platform; create a token from the Inference experiment page
- Rate and usage limits
- Per API key: 3M input and 60K output tokens per 60 seconds; 500M input and 5M output tokens per 24 hours.
- Models
- Qwen/Qwen3.6-35B-A3B-FP8
- Card requirement
- unknown
- Regions
- Global
- After the free allowance
- Requests are rate-limited. Hetzner does not promise continued availability.
- Data and privacy catch
- Check Hetzner's current experiment terms before sending sensitive data.
Do not treat this as guaranteed capacity
The provider describes this access as experimental, rotating, or subject to change. Build a fallback before using it in anything important.