How to Connect Ollama and Make.com: Step-by-Step Guide (2026)

In the evolving landscape of business technology, leveraging artificial intelligence efficiently and securely is paramount. As we look towards 2026, the demand for powerful, privacy-preserving, and cost-effective AI solutions continues to grow. This guide focuses on integrating Ollama, an open-source framework for running large language models (LLMs) locally, with Make.com, a leading visual workflow automation platform. This combination allows businesses to build sophisticated AI-driven workflows that prioritize data security, reduce operational costs, and offer unparalleled customization.

Ollama provides the capability to host and manage LLMs directly on your own infrastructure, giving you full control over your data and models. This eliminates the reliance on third-party cloud APIs for sensitive data processing, addressing growing concerns about data privacy and compliance. Make.com, on the other hand, empowers users to connect thousands of applications and automate complex processes without writing code. By bringing these two platforms together, you can design intelligent automations where a local LLM performs complex natural language processing tasks, and Make.com orchestrates the inputs, outputs, and subsequent actions across your business systems.

Imagine automating customer support responses, summarizing internal documents, or generating personalized marketing content—all powered by an AI model running securely within your own network, seamlessly integrated into your existing workflows. This guide will walk you through the practical steps to achieve this, enabling your organization to harness advanced AI capabilities with enhanced security and efficiency.

What You Need to Get Started

Before diving into the integration process, ensure you have the following components and basic setup in place:

Step-by-Step Guide: Connecting Ollama and Make.com

Step 1: Ensure Ollama is Running and Accessible via a Secure Tunnel

First, confirm that your Ollama server is operational and has the desired model loaded. Ollama typically runs on port 11434 by default (e.g., http://localhost:11434).

  1. Start Ollama: Ensure Ollama is running in the background. You can verify this by running ollama list in your terminal to see downloaded models.
  2. Start Your Tunneling Service: Since Make.com is a cloud service, it cannot directly access your local network. You'll need to use a tunneling service.
    • Example with ngrok: If you have ngrok installed, open a terminal and run: 
      ngrok http 11434

      This will provide you with a public URL (e.g., https://your-random-id.ngrok-free.app). Keep this terminal window open as long as you need the tunnel active.

    Note: Record the public URL provided by your tunneling service. This URL is crucial for Make.com to communicate with your Ollama instance.

Step 2: Create a New Scenario in Make.com

  1. Log in to your Make.com account.
  2. From your dashboard, click on the "Create a new scenario" button. This will open the visual editor.

Step 3: Add a Trigger Module

Every Make.com scenario starts with a trigger. This defines when your automation should run.

  1. Click the large plus icon in the center of the canvas.
  2. Choose a suitable trigger. Common examples include:
    • Webhook: To trigger the scenario from an external application or system (e.g., when new data arrives).
    • Scheduler: To run the scenario at specific intervals (e.g., daily, hourly).
    • Email: To trigger when a new email arrives in your inbox.
    • Google Sheets: To trigger when a new row is added to a spreadsheet.

    For this guide, let's assume you're using a Webhook to receive a prompt.

  3. Configure your chosen trigger module as needed. For a Webhook, you'll copy the provided Webhook URL for later use.

Step 4: Add an HTTP Module to Interact with Ollama

This module will send requests to your local Ollama instance via the public tunnel URL.

  1. Click the "Add another module" button next to your trigger module.
  2. Search for and select the "HTTP" module.
  3. Choose the "Make a request" action.
  4. Configure the HTTP module with the following settings:
    • URL: Paste your secure tunnel URL from Step 1, followed by /api/generate.

      Example: https://your-random-id.ngrok-free.app/api/generate

    • Method: Select POST.
    • Headers:
      • Add a new header:
        • Name: Content-Type
        • Value: application/json
    • Body type: Select Raw.
    • Content type: Select JSON (application/json).
    • Request content: This is the JSON payload that Ollama expects. You'll map data from previous modules here.

      Example structure:

      {
  "model": "llama3",
  "prompt": "{{1.data}}",
  "stream": false
}

      Explanation:

      • "model": "llama3": Replace llama3 with the name of the Ollama model you want to use.
      • "prompt": "{{1.data}}": This is where you dynamically insert the prompt from your trigger module. If your Webhook module (module 1) receives data under a key like "text", you would map it as {{1.text}}. Adjust the mapping based on your trigger's output.
      • "stream": false: Ensures the entire response is returned at once, which is generally easier to process in Make.com.
  5. Click "OK" to save the module configuration.

Step 5: Process the Ollama Response

Ollama will return its response in JSON format. You'll need to extract the relevant text.

  1. Add another module next to the HTTP module.
  2. Search for and select the "JSON" module.
  3. Choose the "Parse JSON" action.
  4. In the "JSON string" field, map the Data output from your HTTP module (e.g., {{2.data}}). This will parse the full JSON response from Ollama.
  5. (Optional) Alternatively, you can directly map specific parts of the HTTP module's output if you know the structure. Ollama's response typically contains a response field within the parsed JSON.

Step 6: Add Subsequent Action Modules

Now that you have the processed text from Ollama, you can use it in other applications.

  1. Add another module.
  2. Choose an action module based on your workflow needs. Examples:
    • Email: Send the generated response via email.
    • Slack: Post the summary to a Slack channel.
    • Google Docs: Append the content to a document.
    • CRM (e.g., Salesforce): Update a record with AI-generated notes.
  3. Configure the action module, mapping the desired output from the JSON Parse module (e.g., {{3.response}}) into the relevant fields.

Step 7: Test and Activate Your Scenario

  1. Click the "Run once" button at the bottom of the Make.com editor to test your scenario.
  2. If using a Webhook trigger, manually send a test request to its URL with a sample prompt.
  3. Review the execution history to ensure all modules ran successfully and the data flowed as expected.
  4. Once confident, toggle the scenario "ON" to activate it.
Ready to set this up? Build this automation free on Make.com.
Start free on Make.com →

Popular Use Cases for Ollama and Make.com Integration

Time Savings Estimate

Integrating Ollama with Make.com significantly reduces manual intervention in tasks involving natural language processing and data orchestration. For a business process that typically requires 15-30 minutes of human effort per instance (e.g., summarizing a document, classifying an inquiry), automation can reduce this to mere seconds. If your team handles 20 such instances daily, this translates to 5-10 hours saved per day. Over a month, this can free up hundreds of hours of employee time, allowing staff to focus on higher-value strategic activities rather than repetitive, time-consuming tasks. The ability to process large volumes of data securely and consistently also minimizes errors and improves overall operational efficiency.

Frequently Asked Questions

Is it secure to expose my local Ollama instance to Make.com?

Yes, it can be secure when implemented correctly. Using a reputable secure tunneling service like ngrok or Cloudflare Tunnel encrypts the traffic between Make.com and your local Ollama instance, protecting it from direct exposure to the public internet. These services often include authentication and access control features. Additionally, ensure your local network where Ollama runs is secure, with appropriate firewall rules and access restrictions, and only expose the necessary port (11434).

Can I use any model with Ollama and Make.com?

Yes, any model that is compatible with and running on your Ollama instance can be utilized with Make.com. Simply ensure the model is downloaded and actively available on your Ollama server. When configuring the Make.com HTTP module, you will specify the exact name of the model in the request content (e.g., "model": "your_model_name").

What are the costs associated with this setup?

The primary costs involve your server infrastructure (hardware, electricity, and internet connectivity) to run Ollama, a Make.com subscription (if your usage exceeds the free tier), and potentially a paid plan for your tunneling service if advanced features or higher bandwidth are needed. A significant benefit is the elimination of per-token API costs typically associated with cloud-based LLM providers, making this a highly cost-effective solution for high-volume or extensive AI processing.

Written by Vangari Sai Sampath, Automation Specialist · Integration Directory · Hyderabad, India