n8n

Requirements

• a running n8n instance;
• a DEVUP AI API key from https://devupai.com/dashboard/api-keys.

Add DEVUP AI as an OpenAI credential in n8n

1. Create a new credential and choose OpenAI (not Azure OpenAI).
2. Paste your DEVUP AI key into API Key and leave Organization ID empty.
3. Replace the Base URL with https://api.devupai.com/v1 and save.

Use Responses API: keep the default

New OpenAI Chat Model nodes have Use Responses API switched on. DEVUP AI works with it on or off, so you can keep the default.

Build an AI Agent workflow with DEVUP AI

1. Add a Chat Trigger and an AI Agent.
2. Attach an OpenAI Chat Model with your DEVUP AI credential, and pick a model under Model → From list.
3. Attach the Calculator tool.
4. Click Open chat below the canvas and send the test prompt.

What is 12 multiplied by 34? Use the calculator tool.

The agent called Calculator and answered 12 multiplied by 34 is 408.

Import the ready-made workflow

Download the workflow, or copy the JSON below and paste it on the n8n canvas with Ctrl+V. Then open OpenAI Chat Model and select your DEVUP AI credential.

{
  "name": "DEVUP AI Agent with Calculator",
  "nodes": [
    {
      "parameters": {
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "typeVersion": 1.5,
      "position": [
        208,
        320
      ],
      "id": "6a8f7c9e-a488-46fe-a577-e7505c8cf081",
      "name": "When chat message received",
      "webhookId": "63355992-9e13-41df-87d8-d96af86012f6"
    },
    {
      "parameters": {
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.agent",
      "typeVersion": 3.1,
      "position": [
        480,
        320
      ],
      "id": "b2d5191a-981c-4d93-909a-7d31dc058afd",
      "name": "AI Agent"
    },
    {
      "parameters": {
        "model": {
          "__rl": true,
          "mode": "list",
          "value": "deepseek-ai/DeepSeek-V4-Pro"
        },
        "responsesApiEnabled": false,
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "typeVersion": 1.3,
      "position": [
        368,
        544
      ],
      "id": "d57704e7-47b2-4857-9fc4-1d01b4d3eb83",
      "name": "OpenAI Chat Model"
    },
    {
      "parameters": {},
      "type": "@n8n/n8n-nodes-langchain.toolCalculator",
      "typeVersion": 1,
      "position": [
        608,
        544
      ],
      "id": "47c9e445-515a-480e-9a49-e69ac2a1f608",
      "name": "Calculator"
    }
  ],
  "pinData": {},
  "connections": {
    "When chat message received": {
      "main": [
        [
          {
            "node": "AI Agent",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "OpenAI Chat Model": {
      "ai_languageModel": [
        [
          {
            "node": "AI Agent",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Calculator": {
      "ai_tool": [
        [
          {
            "node": "AI Agent",
            "type": "ai_tool",
            "index": 0
          }
        ]
      ]
    }
  },
  "settings": {
    "executionOrder": "v1",
    "binaryMode": "separate"
  }
}

The OpenAI node's Message a model action

The Message a model action of the OpenAI node works with DEVUP AI chat models. Set Model to By ID and enter a catalog ID, such as deepseek-ai/DeepSeek-V4-Pro.

Use DEVUP AI for the n8n Assistant

In the Assistant's Connect a model step, choose Self-hosted or OpenAI-compatible endpoint, enter https://api.devupai.com/v1 and your key, and type an OpenAI model such as openai/gpt-6-luna. The Assistant's next step asks for a code sandbox, which is a separate service run by n8n or Daytona, not DEVUP AI.

Choosing models

Use the exact catalog ID. Tool calling needs a model whose catalog entry supports it. See the Models catalog and Tool Calling guide.

Troubleshooting

• Couldn't connect with these settings:The Base URL still points at the default endpoint. Set it to https://api.devupai.com/v1 and save again.
• The chat echoes your message back as JSON:The chat was opened from inside the Chat Trigger, which runs only that node. Use Open chat below the canvas to run the whole workflow.
• The Assistant's sandbox step returns an unexpected response:The sandbox is a separate service. Enter the URL and key of an n8n Sandbox or Daytona there, not DEVUP AI.

FAQ

Related integrations

• LangChain: Chains and agents with LangChain.
• OpenAI SDK: OpenAI SDKs pointed at DEVUP AI.
• Node.js SDK: The official DEVUP AI SDK for Node.js.
• Python: The official OpenAI Python SDK on DEVUP AI.