Hetzner Experiments Inference API
Free while the provider's public experiment remains active.
https://inference.hetzner.com/api/v1Models mentioned
Qwen/Qwen3.6-35B-A3B-FP8Qwen3.8-27BEquivalent paid value
How this was valued: Continuous saturation of the documented per-API-key 60-second window (10 requests, 4M input tokens, 100k output tokens) with each request's total tokens capped by the served models' documented 262,144-token context, priced for the costlier served model, Qwen3.8-27B, at the cheapest current OpenRouter listing for the exact model because neither Hetzner (free while experimental) nor Alibaba Cloud Model Studio's international catalog publishes a paid rate for these exact models. The context-derived per-request ceiling (10 x 262,144 = 2,621,440 tokens/minute) is tighter than the 4M input-TPM limit and was used. Assumes uninterrupted saturation of one API key with no latency, concurrency, availability, or fair-use loss; the value is a paid-equivalent ceiling, not expected usage.
Limits and terms
- Window Seconds
- 60
- Requests
- 10
- Input Tokens
- 4,000,000
- Output Tokens
- 100,000
- Scope
- per api key
What happens to your prompts?
Hetzner documents no ordinary prompt/output retention, no content training right, and EU-hosted processing.
- Plan Scope
- Hetzner Experiments Inference API, a best-effort experimental service for existing Hetzner customers.
- Prompt Retention
- Hetzner states it does not retain prompt or response content after processing unless legally compelled.
- Response Retention
- Same non-retention statement as prompts.
- Ordinary Logging
- Usage metadata is retained for operation and enforcement without prompt or response content.
- Model Training
- No retained inference content is available for ordinary model training, and no training right is disclosed for the experimental API.
- Product Improvement
- Service telemetry may support operation of the experiment; the product documentation does not authorize content-based improvement.
- Human Or Operator Access
- Transient access necessary to process a request is possible; retained content access is limited to a legal-compulsion exception.
- Subprocessors And Routing
- Hetzner provides the hosted inference infrastructure in the EU under its terms and DPA; model licenses can separately restrict use of outputs.
- Deletion Controls
- Prompt and response content is not ordinarily retained. Account and personal-data rights are available through Hetzner's privacy and DPA procedures.
- Caveat
- This is a non-production experiment without an availability commitment. The legal-compulsion exception and retained usage metadata mean it is not a universal zero-data promise.
Governing documents
- Terms And Conditions
- https://www.hetzner.com/legal/terms-and-conditions
- Privacy Policy
- https://www.hetzner.com/legal/privacy-policy
- Data Processing Agreement
- https://www.hetzner.com/AV/DPA_en.pdf
Eligibility
- Account Required
- Yes
- Hetzner Customer Required
- Yes
- Payment For Inference Required
- No
Primary sources
Before you build with Hetzner Experiments Inference API
Read the classification narrowly
This record has current first-party evidence for some zero-cost hosted inference. The label describes the bounded offer supported by evidence on 2026-09-19; it does not guarantee permanence, production suitability, uptime, latency, model quality, or access from every region.
Resolve the live model route
This snapshot records 2 model IDs. Match the exact ID against the provider's live catalog before using it in code; zero-price routes and aliases can rotate while an older documentation page remains online. The recorded base endpoint is https://inference.hetzner.com/api/v1, but the provider's current API reference remains authoritative for paths, authentication, and request shape.
Confirm account and billing boundaries
An account is required, so public catalog evidence cannot by itself prove the limits, balance, or billing behavior attached to your workspace. The reviewed evidence explicitly says no payment card is required. Re-check the signup flow because eligibility and billing controls can change after the snapshot. The equivalent paid value shown here prices a documented allowance where possible; it is not cash, guaranteed savings, or protection from overage.
Re-read the prompt policy for your plan
This record classifies the reviewed handling as “Private”: Hetzner documents no ordinary prompt/output retention, no content training right, and EU-hosted processing. Provider policies can distinguish free, paid, enterprise, opted-in, and feature-specific traffic, so verify the governing terms for the exact account and route that will receive your data.
Follow the evidence, then re-check it
The record links 11 first-party sources covering the offer, catalog, limits, pricing, terms, privacy, or adoption evidence available to the audit. Prefer the newest governing document or live catalog when sources disagree, and submit a correction when a provider changes a material term.
Match the quota to the workload shape
Translate the published allowance into the traffic pattern you actually expect instead of comparing headline totals alone. A daily token pool can look generous while a low requests-per-minute or concurrency ceiling blocks interactive bursts; a high request limit can still fail a long-context job when input, output, or per-request tokens are capped. Separate prompt tokens from generated tokens, include retries and tool calls, and test the largest realistic payload. If the provider documents more than one limit window, the tightest window at your peak load is the practical ceiling. Research, experimental, and community access can also carry eligibility or fair-use constraints that cannot be modeled as a simple number.
Plan fallback without changing the rules
A fallback should preserve more than API syntax. Confirm that the substitute route supports the required modality, context length, streaming behavior, structured output, tools, and safety controls, then compare its prompt-retention and training terms. An OpenAI-compatible request shape does not make providers operationally or contractually equivalent. Decide which errors may trigger a retry, cap retry storms, and prevent an exhausted free route from silently switching to a billable model. If deterministic output matters, record the model revision and sampling settings; rotating aliases and free-model pools can change behavior even when the endpoint remains available.
Monitor the offer as a dependency
Capture the model ID, response model field, rate-limit headers, usage fields, latency, HTTP status, and any provider request identifier for each test. Watch for authorization failures, quota exhaustion, catalog removal, policy revisions, and dashboard balance changes as separate failure modes. Re-check the provider’s live catalog and governing pages on a schedule proportionate to the workload’s importance, and keep the dated sources that supported your decision. Free capacity is especially suitable for prototypes, evaluations, fallbacks, and bounded workloads when the application can tolerate change; a production dependency still needs observability, an exit path, and an owner responsible for re-verification.
Test one complete request before scaling
Start with the smallest permitted request using the exact credential, model ID, endpoint, and account type you intend to deploy. Record the HTTP status, response headers, usage fields, latency, and any dashboard balance change. Then exercise the failure path: an invalid model, an exhausted quota, or a rate-limit response should fail clearly without silently switching to a paid route. If the provider supports streaming or tool calls, validate those features separately because a free model can expose a narrower capability set than its paid counterpart. Keep a budget ceiling outside the application whenever billing is possible, and avoid sending sensitive data until the observed route matches the reviewed data agreement.