Run Any Ollama AI Model with n8n 🦙🧠

Tired of paying for AI models? 😩 This is how to run them locally with n8n! 🤯

This is your one-stop shop for integrating powerful, open-source AI models like Llama 3.1 directly into your n8n workflows.

No more relying on expensive cloud services! 🚀

Setting the Stage 🎬

What you’ll need:

• A basic understanding of n8n (it’s like Zapier or Make, but better! 😉)
• A server with Docker installed (I recommend Coolify – super easy to use!)

Why this matters:

• Cost savings: 💰 Stop paying for every API call and use powerful models for free.
• Privacy: 🔐 Keep your data on your own server, under your control.
• Customization: 🔧 Fine-tune models to your exact needs.

Installing Ollama: Your AI Powerhouse 🔌

1. Get the Docker image:

   docker pull ollama/ollama

2. Run Ollama (CPU Only):

   docker run -d -p 11434:11434 ollama/ollama

• For Nvidia GPUs: Check Ollama’s documentation for specific instructions (https://hub.docker.com/r/ollama/ollama).

Bridging the Gap: Connecting Ollama and n8n 🔗

Think of it like this: you need to get Ollama and n8n talking on the same network. 🗣️

1. Find your n8n network:

   docker network ls

• Look for the network your n8n container is using (you can find this in Coolify or Portainer too).

1. Find your Ollama container ID:

   docker ps

3. Connect Ollama to your n8n network:

   docker network connect <YOUR_N8N_NETWORK_NAME> <YOUR_OLLAMA_CONTAINER_ID>

• Replace <YOUR_N8N_NETWORK_NAME> and <YOUR_OLLAMA_CONTAINER_ID> with the values you found.

1. Restart your containers: This ensures the connection takes effect.

Installing Your Desired Model 🧠

1. Access your Ollama container:

   docker exec -it <YOUR_OLLAMA_CONTAINER_ID> ollama run llama3.1 

• Replace <YOUR_OLLAMA_CONTAINER_ID> with your Ollama container ID.
• Replace llama3.1 with your desired model (e.g., mixtral, llama2).

Unleashing the Power: Using Ollama in n8n 🚀

1. Get your Ollama container’s IP address:

   docker inspect <YOUR_OLLAMA_CONTAINER_ID>

• Look for the IPAddress field in the output.

1. In n8n, add an “AI Agent” node.

2. Select “Ollama Chat” as your chat model.

3. Enter your Ollama Base URL:

• Use the format: http://<YOUR_OLLAMA_CONTAINER_IP>:11434

You did it! 🎉

You’ve unlocked a world of AI possibilities with locally run Ollama models in your n8n workflows! Now go build something amazing! 🤖

Resources 🧰

• Ollama Docker Hub: https://hub.docker.com/r/ollama/ollama – Get Ollama and explore different models.
• Ollama Website: https://ollama.com/ – Learn more about Ollama and its capabilities.
• Coolify: https://coolify.io/self-hosted – Simplify your Docker experience with this awesome tool.
• Portainer: https://portainer.com – Another great GUI for managing Docker containers.
• n8n: https://n8n.io/ – The ultimate workflow automation tool!