European inference infrastructure
Fast open-model inference, served in Europe.
OpenAI-compatible access to open models, with inference served from Belgium and the Netherlands.
Operational trust
Independent EU inference.
Vaks is designed for EU-resident routing, with inference requests handled in Belgium and the Netherlands.
No prompt retention
Inference prompts are not logged.
No response retention
Generated responses are not logged.
No training
Customer data is not used to train models.
EU processing
Inference is served from Belgium and the Netherlands.
Open model · Text output
Qwen3.8-27B
Open-model inference with the API capabilities platform teams expect.
- —OpenAI-compatible API
- —Streaming responses
- —Structured JSON output
- —Tool calling
Contact