Skip to main content

Configure an n8n agent with Clovis

This page explains how to connect n8n to Clovis via the OpenAI-compatible connector (Clovis base URL), and then how to use it in an AI Agent.

Prerequisites​

Access to n8n (Cloud or self-hosted)
A Clovis API key
The Clovis API URL (OpenAI-compatible)
Example: https://llm-gateway.clovis-ai.fr/v1
The model name / model group to use
Example: ClovisLLM

Step 1 — Foundation​

1- Create a workflow

Create a workflow

2- Create an agent

Create an agent

Step 2 — Configure the “OpenAI Model” node​

1- Add an OpenAI Chat Model node to your workflow.

Add an OpenAI node

2- In Credential to connect with, select Create new credential.

Create new credential

3- Create a new OpenAI credential:

API Key: your Clovis key (sk-XXXXXXXXXX)
Base URL: https://llm-gateway.clovis-ai.fr/v1
Organization ID: optional (leave empty in most cases)

OpenAI credential

4- In Model, select or enter the Clovis model (e.g., ClovisLLM/gpt-oss-120b).

OpenAI Model

Step 3 — Connect the model to an AI Agent​

  1. Connect the OpenAI Model node to the agent’s Chat Model port.
  2. (Optional) Add Memory if you want to keep conversational context across executions.
  3. (Optional) Add Tools if you want to equip the agent (e.g., HTTP Request, Google Drive, etc.).

Connect the model

Step 4 — Test with a simple example (summary)​

In your AI Agent, you can use instructions such as:

  • System / Instructions (recommended) “You are a summarization assistant. Summarize in a maximum of 5 bullet points, with a neutral tone. If information is missing, mention it.”
  • User message (input data) “Here is a text: {{$json.text}}. Provide a summary.”

Then:

  1. Click Execute workflow
  2. Check the output of the AI Agent node (generated summary)

This workflow is enough to validate the n8n ↔ Clovis connectivity and start generating summaries, classifications, reformulations, etc.