n8n
n8n's OpenAI Chat Model node works with Cloakray through an OpenAI credential with a different base URL.
Add the credential
In n8n, create a new OpenAI credential:
| Field | Value |
|---|---|
| API Key | Your Cloakray key |
| Organization ID | Leave blank |
| Base URL | https://api.cloakray.com/v1 |
Keep the key in the credential rather than in a workflow expression or prompt.
Build a workflow
- Add an On chat message trigger.
- Connect it to an AI Agent node.
- Under the agent's Chat Model connector, add OpenAI Chat Model and select the credential.
- Turn Use Responses API off, set the model selector to By ID, and enter a model ID from the catalogue, such as
gpt-5.5orclaude-sonnet-4-6. - Click Open chat and send a message.
For agents that call tools, pick a model with tool support. The other OpenAI nodes — embeddings, image generation — take the same credential.
Troubleshooting
- Authentication fails — check the key and that the base URL ends in
/v1. - The model is missing from the list — choose By ID and type the model ID.
- Requests fail — confirm Use Responses API is off, and check the failing step's error message.