---
title: "Greenference free inference"
description: "A possible free offer exists, but a decisive term still needs verification."
canonical_url: "https://freeinferencing.com/provider/greenference/"
md_url: "https://freeinferencing.com/provider/greenference.md"
last_updated: "2026-10-06"
---

# Greenference

> A possible free offer exists, but a decisive term still needs verification.

## Classification

- **Directory:** Watchlist
- **Free access:** Free model routes
- **Status:** Needs Verification
- **Confidence:** Medium
- **Payment card:** No
- **Account:** Required
- **Equivalent paid value:** Not quantifiable
- **API endpoint:** `https://llm.eu.greenference.com/v1`

## Models mentioned

- `greenference/gemma-4-26b-a4b-it`
- `greenference/gpt-oss-20b`
- `greenference/qwen3-14b`

## Limits and terms

```yaml
requests_per_minute: 60
quota_caveat: Flat per-account request limit; no public free-specific billing quota.
```

## What happens to your prompts?

**Partially private.** Privacy protections are incomplete, conditional, or not fully documented.

Privacy is audited separately from price. Review the current governing terms before sending sensitive or regulated data.

## Data governance and agreements

```yaml
data_governance:
  review_status: partial
  plan_scope: Direct business API; marketplace users contract with their router.
  prompt_retention: DPA forbids persistent content retention; volatile prefix
    caching may outlast a request.
  response_retention: No durable completion retention under DPA.
  ordinary_logging: Content-free request metadata90days; security/IP12months; billing10years.
  model_training: DPA excludes training, evaluation, datasets and all secondary
    content use for every tier.
  product_improvement: Content-based improvement prohibited.
  human_or_operator_access: DPA forbids inspecting content; transient plaintext is
    processed on EU rented hardware.
  subprocessors_and_routing: Vast.ai and Vultr receive plaintext in memory on EU
    hosts; US-established operators and remote-access transfer safeguards are
    disclosed.
  deletion_controls: Data-subject and metadata deletion rights; no stored content to delete.
  conflict: Privacy page contains unresolved draft placeholders and older
    document-precedence/change-notice language inconsistent with the newer
    Terms/DPA. Conservative partial review retained.
agreements:
  terms_of_service: https://greenference.com/legal/terms
  data_processing_addendum: https://greenference.com/legal/dpa
  privacy_policy: https://greenference.com/legal/privacy
  acceptable_use_policy: https://greenference.com/legal/aup
```

## Eligibility

```yaml
account_required: true
business_registration_required: true
professional_use_only: true
payment_method_required: not_verified
```

## API compatibility and modalities

- **Compatibility:** OpenAI Compatible
- **Modalities:** Text generation


## Before you build with Greenference

### Read the classification narrowly

This record describes the bounded offer supported by evidence on 2026-10-06; it does not guarantee permanence, production suitability, uptime, latency, model quality, or access from every account and region. “Free model routes” should be interpreted together with the Watchlist directory placement, Medium confidence, No payment-card status, and the plan-specific sources below.

### Resolve the live model route

The snapshot records 3 model IDs. Match the exact ID against the provider's current catalog before using it in code because zero-price routes, aliases, context limits, and feature support can rotate while an older documentation page remains online.

### Match the quota to the workload shape

Translate the allowance into peak requests per minute, input and output tokens, concurrency, retries, and every daily or monthly ceiling that applies to the intended account. The first limit reached by the workload is the practical ceiling. A large token pool can still fail interactive bursts, and a high request limit can still fail long-context work. Include tool calls and retry traffic, test the largest realistic payload, and treat research, experimental, and community access as conditional on their eligibility and fair-use terms.

### Preserve billing and privacy boundaries

The recorded payment-card requirement is No, and account access is required. Confirm both in the live signup flow, set provider-side budgets when charges are possible, and observe the balance or usage fields after a complete request. The prompt-handling classification is **Partially private** because Privacy protections are incomplete, conditional, or not fully documented. Re-read the terms for the exact route and plan before sending sensitive, regulated, or proprietary content.

### Follow the evidence, then re-check it

This record links 9 first-party sources covering the offer, catalog, limits, pricing, terms, privacy, or adoption evidence available to the audit. Prefer the newest and most specific governing document or live catalog when sources disagree. A dated finding can become stale even when the page remains online, so retain the source and snapshot that supported the decision and submit a correction when a provider changes a material term.

### Plan fallback without policy drift

A fallback should preserve modality, context length, streaming, structured output, tools, safety controls, and data terms, not only API syntax. Decide which errors may retry, cap retry storms, and prevent an exhausted free route from silently switching to a billable model. Rotating aliases and free pools can change behavior without changing the endpoint, so retain the response model field and test at least one substitute route before the primary offer becomes unavailable.

### Monitor the offer as a dependency

Capture the model ID, response model, rate-limit headers, usage fields, latency, HTTP status, and provider request identifier. Watch authorization failure, quota exhaustion, catalog removal, policy revision, and balance movement as separate failure modes. Re-check the live catalog and governing sources on a schedule proportionate to the workload's importance, and keep an owner and exit path for any production dependency on volatile free capacity.

### Test one complete request before scaling

Start with the smallest permitted request using the exact credential, model ID, endpoint, region, and account type intended for deployment. Record the status, headers, usage fields, response model, latency, and dashboard balance movement. Then exercise an invalid model, quota exhaustion, or rate limit so failure is explicit and cannot silently switch to a paid route. Validate streaming, structured output, and tool calls separately because a free model can expose fewer features than its paid counterpart. Keep a budget ceiling outside the application whenever billing is possible, and do not send sensitive data until the observed route matches the reviewed agreement.

## Primary sources

- [Greenference model pricing page](https://greenference.com/models): Official Pricing
- [Greenference public models JSON](https://llm.eu.greenference.com/v1/models): Live Catalog
- [Greenference API](https://greenference.com/docs): Official Docs
- [Greenference direct API Terms](https://greenference.com/legal/terms): Official Terms
- [Greenference rate limits](https://greenference.com/docs/rate-limits): Official Docs
- [Greenference DPA](https://greenference.com/legal/dpa): Official Dpa
- [Greenference Privacy Policy](https://greenference.com/legal/privacy): Official Privacy
- [Greenference Acceptable Use Policy](https://greenference.com/legal/aup): Official Aup
- [Greenference rented hardware and gateway providers](https://greenference.com/legal/subprocessors): Official Subprocessors

## Sitemap

See the full [semantic sitemap](/sitemap.md) for every page and markdown mirror.