French inference host for small open-weight models, served from GPU capacity in the EU.
Greenference SAS is a French company, registered in Lyon, that serves small open-weight models through an OpenAI-compatible API. It runs on GPU capacity rented inside the European Union, behind a request router in Amsterdam. On Opper it serves the Qwen3, Qwen3.5, Qwen3.6 and Qwen3.8 families alongside Gemma 4, GLM-4.7-Flash, gpt-oss-20b and Llama 3.1 8B. A good pick for inexpensive EU-hosted inference on 8B to 32B class models.
Every model we route through Greenference. Compare residency, zero data retention, training posture and price at a glance, with full data-handling detail per route below.
| Model | Region | Zero data retention | Training | Context | Input | Output |
|---|---|---|---|---|---|---|
| EU | No | 262K | $0.05 | $0.90 | ||
| EU | No | 16K | $0.15 | $1.00 | ||
| EU | No | 203K | $0.03 | $0.20 | ||
| EU | No | 16K | $0.04 | $0.07 | ||
| EU | No | 33K | $0.0090 | $0.04 | ||
| EU | No | 16K | $0.01 | $0.02 | ||
| EU | No | 131K | $0.07 | $0.35 | ||
| EU | No | 41K | $0.10 | $0.28 | ||
| EU | No | 33K | $0.45 | $0.45 | ||
| EU | No | 16K | — | — | ||
| EU | No | 16K | $0.10 | $0.28 |
Greenference hosts on 1 route. Each route has its own privacy posture, residency, and GDPR terms. Postures are maintained by Opper with a last-verification timestamp.
Zero data retention: nothing is logged, held for abuse monitoring or used for training. No training on customer data. EU; DPA available.
