Choose n8n in Guided Connection
Clyvel includes an n8n connection path that walks through the gateway URL, workload credential and test request without placing provider secrets into the chat history.
All integrations
N8N INTEGRATION
n8n OpenAI Monitoring, Cost Visibility and AI Gateway Integration
Automation workflows can create many model calls without making their operational cost or failure pattern obvious. Connecting n8n model traffic through Clyvel gives the workflow a shared production layer where requests can be attributed, inspected and compared across models and providers.
WHEN THIS FITS
A practical production use case.
n8n workflows that call OpenAI-compatible model endpoints from production automations.
Teams that need to understand which workflow is driving tokens, latency or AI spend.
Operators who want model failures visible outside individual n8n execution logs.
CONNECTION PATH
Connect the model request layer, then verify evidence.
Route the AI request through Clyvel
Configure the relevant OpenAI-compatible request in n8n to use the Clyvel Gateway endpoint and the Clyvel key assigned to the workload.
Run the workflow and confirm the request
Execute the workflow once and verify that Clyvel captured the request before expanding the integration to additional automations.
01
See AI request activity outside isolated workflow executions.
02
Attribute token and estimated provider cost to production automation traffic.
03
Investigate model errors and latency with the same operational context used by other AI applications.
FAQ
Integration questions.
Can Clyvel monitor AI requests created by n8n workflows?
Yes when the workflow routes its supported model request through the Clyvel Gateway. The request can then appear with operational telemetry such as model, provider, latency, status, tokens and cost evidence.
Does Clyvel replace n8n execution history?
No. n8n remains the workflow runtime. Clyvel focuses on the AI request layer so model traffic can be observed and controlled consistently alongside other applications.
NEXT STEP
Connect one workload before expanding the rollout.
Start with a controlled request, confirm that operational evidence is trustworthy, and only then move more production traffic or add routing and reliability policy.