How to Connect n8n to a Local LLM: Building Private AI Automation with Ollama

Description and use

n8n connects to a local LLM Ollama via an internal Docker network, enabling automation without sending data to external services.

Problem

Running AI-powered workflows on private infrastructure while ensuring sensitive data never leaves the internal network.

Outcome

n8n workflows can now invoke a locally hosted LLM without external API calls, keeping data within the internal network and reducing privacy risks.

Workflow

• Deploy n8n and Ollama containers with Docker Compose, ensuring Ollama is on a private network.
• Download required model (e.g., qwen2.5:1.5b) into Ollama before isolating the network.
• Configure n8n HTTP Request node to POST to http://ollama:11434/api/chat with JSON payload.
• Disable streaming in Ollama to receive full response in one JSON.
• Parse response message.content to extract summary and category.
• Store structured output in desired destination (e.g., database, email).
• Verify connectivity with GET http://ollama:11434/api/tags and POST /api/chat tests.
• Implement network isolation: Ollama only on internal network, n8n on frontend and inference networks.
• Add authentication gateway if needed, otherwise rely on network isolation.
• Monitor logs, execution history, and backups to ensure no sensitive data leaks.

Evidence

Deploy Ollama alongside n8n, configure an HTTP request, and within minutes, your workflow can summarize support tickets, classify requests, or process internal documents without sending their contents to an external AI provider.

Source

The Ollama container is connected only to the internal inference network. Its API port is not published to the host or public Internet.

Source

Authentication requirements depend on how Ollama is deployed. By default, Ollama's local API does not require an API key.

Source