I have created an n8n workflow that solves model selection, routing, and data persistence via a unified API (i.e. no more individual integrations or manual adjustments when switching models).
In short: the foundation for the European "Open Router" in n8n is therefore laid...
🧩 What the workflow does:
* Connects n8n with Requesty via an OpenAI-compatible interface * Supports both fixed model usage and dynamic selection
* Stores selected models in Google Sheets
* Uses the chosen model for response processing
* Technically clean separation of selection, control, and execution
⚙️ Two setup variants:
1. Directly via OpenAI node:
* Model selectable directly in the node
* Easy to configure
2. Dynamically with HTTP workflow:
* Model list is retrieved via API
* Selection in the form trigger (dropdown)
* Storage in Google Sheets
* Automated further processing with the selected model
🛡️ GDPR-compliant with Requesty:
* Unified API for multiple models
* Provider with DPA and based in the European legal area
* No direct communication with US APIs required
* Setup reduces legal complexity and technical fragmentation
Follow-up questions or ideas for further development are, as always, welcome 🤝
PS: Attached is the workflow. It will also appear on n8n in a few days, provided the submission is accepted...