Agents
n8n
Last updated March 3, 2026
Connect n8n workflow automation to Cencori for AI-powered workflows with model routing and monitoring.
n8n
n8n is a low-code workflow automation platform with 300+ integrations. When connected to Cencori, all AI calls in your workflows route through the Cencori gateway — giving you model switching, billing, shadow mode, and audit logging without changing your workflows.
[!TIP] n8n has two connection modes: API Control Plane (one-click install, recommended) and Manual (import templates yourself). Choose API mode for the smoothest experience.
Architecture
Quick Start
1. Deploy from the Marketplace
Navigate to Agents → Agent Marketplace in your project dashboard and click Deploy Agent on the n8n card.
2. Generate an API Key
Go to the agent's Configuration tab and click Generate Key. Copy the key — you'll only see it once.
3. Install n8n
npx n8n start[!NOTE] Or use Docker:
docker run -it --rm -p 5678:5678 n8nio/n8n. See n8n docs for all install methods.
4. Configure n8n Credentials
In your n8n instance:
- Go to Settings → Credentials → Add Credential → OpenAI API
- Set Base URL to:
https://cencori.com/api/v1 - Set API Key to your
cake_key from Step 2 - Save the credential
5. Use in Workflows
Add an OpenAI Chat Model node to any workflow:
- Select the credential you just created
- Set the Model to the model configured in your Cencori dashboard (e.g.,
gpt-4o) - All requests now route through Cencori
6. Connect via Dashboard (Optional)
For full control plane features (one-click workflow install, publish/unpublish, health monitoring), connect your n8n instance from the agent's Settings tab:
- Enter your n8n Base URL (e.g.,
https://your-n8n.example.com) - Enter your n8n API Key (from n8n Settings → API)
- Click Connect
Capabilities
| Feature | Description |
|---|---|
| Visual Workflows | Drag-and-drop workflow builder |
| 300+ Integrations | Slack, GitHub, Notion, Google Sheets, and more |
| LangChain Nodes | Advanced AI chains with memory and tools |
| Webhook Triggers | Event-driven automation |
| Scheduled Tasks | Cron-based recurring workflows |
Configuration
Model Selection
Choose any supported model from the Cencori Configuration tab. n8n picks up the change on the next workflow execution — no n8n restart needed.
Shadow Mode
When enabled, any tool calls returned by the LLM are held as pending in the dashboard. You approve or reject them before n8n executes the next step.
Troubleshooting
n8n credential test fails
- Verify the Base URL is
https://cencori.com/api/v1(not/v1/chat/completions) - Check that your
cake_key is correct and not revoked - Ensure your Cencori project has remaining spend allowance
Workflows not using Cencori
- Confirm the OpenAI credential in the workflow node points to Cencori, not
api.openai.com - Check the workflow's execution logs for routing details