Hi n8n community! 
I’m Rogier from NL AI Gateway — a sovereign, Dutch-hosted LLM API that’s fully OpenAI-compatible. If you’re building AI workflows in n8n but have concerns about sending customer data to US-based providers, this might be for you.
Why we built this
Most n8n users default to OpenAI for AI tasks. But for European businesses handling HR data, medical records, legal documents, or customer PII, sending that data to US servers creates GDPR compliance headaches. We wanted an alternative that:
Runs on Dutch infrastructure (no US/EEA data export)
Has a zero data retention policy — prompts & responses are never logged
Has a no-model-training guarantee — your data is never used to train AI models
Is fully OpenAI-compatible — drop-in replacement, just change the base URL
How to use it in n8n (takes 20 seconds)
- Add an OpenAI Chat Model node (or any HTTP Request node)
- Create new credentials with your API key — get one at nl-ai-gateway.com
- Check Custom API URL and enter:
https://gateway.nl-ai-gateway.com/v1
That’s it. Your AI workflows now run on sovereign EU infrastructure.
What’s under the hood
- Qwen 3.6 35B model (open-weights, self-hosted on Dutch servers)
- Full streaming support (SSE)
- Tools & function calling — works with n8n’s tool-use patterns
- Web search capability (via SearXNG)
- Anthropic Messages API also supported (
/v1/messages) - Subscription tiers from solo to enterprise
Official n8n Community Node — just published! 
We also published an installable n8n Community Node: n8n-nodes-nl-ai-gateway. Install it via Settings → Community Nodes — search for “NL AI Gateway” or “Dutch AI”, click Install, and you get dedicated nodes for Chat Completion, Summarize Text, and List Models — no manual Base URL configuration needed.
Try it for free
I’d love to get feedback from this community. Sign up and try 14 days for free.
Questions?
Happy to answer anything about the infrastructure, model performance, compliance, or n8n setup. Drop a comment below!
