How to Use Make.com with AI Tools for Workflow Automation

Step by step

1. Create a new scenario in Make.com

   Log into your Make.com dashboard and click Create a new scenario. Choose a trigger module from the left panel — this could be Gmail for email triggers, Webhooks for external data, or Schedule for time-based automation. Drag your chosen trigger onto the canvas.

   Note: Start with simple triggers like Form submissions or New file in Google Drive before moving to complex webhook integrations.

2. Add an AI module to your workflow

   Click the plus icon after your trigger module. Search for your preferred AI service — OpenAI, Anthropic Claude, or Google AI. Select the specific action like Create a Chat Completion for OpenAI or Create a Message for Claude. Drag the module onto your canvas and connect it to the trigger.

   Note: OpenAI offers the most Make.com integrations, but Claude excels at longer text analysis and Gemini works best for Google Workspace integration.

3. Configure your AI module connection

   Click on the AI module and select Create a connection. Enter your API key from your AI service account — find this in OpenAI Platform settings, Anthropic Console, or Google AI Studio. Name your connection and test it to verify the authentication works correctly.

   Note: Store multiple API keys as separate connections to switch between different AI models or accounts within the same scenario.

4. Set up your AI prompt and parameters

   Configure the Model field with your preferred AI model like gpt-4, claude-3-sonnet, or gemini-pro. In the Messages field, structure your prompt using dynamic data from previous modules. Set Temperature between 0.1 for consistent responses or 0.7 for creative outputs. Adjust Max Tokens to control response length.

   Note: Use Make.com's mapping feature to insert trigger data directly into prompts — map email content, form responses, or file data as variables in your AI requests.

5. Add output actions for AI responses

   Connect modules after your AI module to handle the generated content. Common actions include Gmail Send an Email, Google Sheets Add a Row, Slack Send a Message, or Airtable Create a Record. Map the AI response content to the appropriate fields in your output module.

   Note: Parse AI responses using Make.com's text functions if you need specific data formats — extract JSON from AI outputs or split responses into multiple fields.

6. Test and refine your AI workflow

   Click Run once to test your complete scenario with sample data. Review each module's output to verify the AI responses meet your requirements. Adjust prompts, model parameters, or add filtering modules if needed. Use the scenario history to debug any errors in the AI processing chain.

   Note: Enable data stores between runs to maintain context across multiple AI interactions — useful for chatbots or progressive data analysis workflows.

7. Schedule and activate your AI automation

   Set your scenario schedule in the trigger module — choose immediate for real-time processing or set intervals for batch processing. Click the toggle switch to activate your scenario. Monitor the execution history for successful AI API calls and any rate limit issues.

   Note: Set up error handling with alternative AI models as fallbacks — if OpenAI hits rate limits, automatically retry with Claude or another service.