n8n is an open-source workflow automation platform with strong AI integration capabilities — connecting LLMs, APIs, and data sources into automated pipelines without extensive custom coding. This guide covers self-hosting it on a VPS.
What n8n Enables
Visual workflow building connecting triggers (webhooks, schedules, form submissions) to actions (API calls, LLM prompts, database operations) — useful for building AI-powered automation (content generation, data processing, chatbot logic) without writing a full custom application.
Step 1 — Install via Docker
docker run -d --name n8n -p 5678:5678 -v n8n_data:/home/node/.n8n n8nio/n8nSee How to Install Docker on AlmaLinux/Rocky Linux or general Docker setup guides for the base Docker installation if needed.
Step 2 — Access the n8n Web Interface
http://YOUR_SERVER_IP:5678Step 3 — Set Up Reverse Proxy with HTTPS
See How to Set Up Nginx as a Reverse Proxy for Node.js Apps for the general pattern — never expose n8n directly on its default port without a properly configured reverse proxy providing HTTPS, given the sensitive credentials/API keys workflows typically involve.
Step 4 — Connect an LLM Node
n8n has built-in nodes for connecting to various LLM providers (both API-based and self-hosted, including compatibility with local models via LocalAI, see How to Install LocalAI as an OpenAI-Compatible API Alternative) — configure credentials and the specific prompt/model within the workflow builder.
Building a Simple AI Automation Example
Trigger: Webhook receives a customer support email
-> Node: Extract email content
-> Node: LLM analyzes sentiment and category
-> Node: Route to appropriate team based on classification
-> Node: Send Slack notificationA representative workflow combining a trigger, AI processing, and automated action — the kind of automation n8n's visual builder makes accessible without custom application development.
Connecting to Your Own Self-Hosted AI Services
See How to Install Ollama and Run Local LLMs on a VPS — n8n can call your own self-hosted LLM endpoints via HTTP request nodes, letting you build automation entirely on self-hosted infrastructure without third-party AI API dependency.
Setting Up Persistent Storage for Workflow Data
docker run -d --name n8n -p 5678:5678 -e DB_TYPE=postgresdb -e DB_POSTGRESDB_HOST=your-db-host -v n8n_data:/home/node/.n8n n8nio/n8nFor production use, configure n8n to use PostgreSQL rather than its default SQLite storage, providing better reliability and backup capability for your workflow definitions and execution history.
Securing Workflow Credentials
See How to Manage Secrets in a CI/CD Pipeline for the general credential management principle — n8n stores credentials for connected services (API keys, database access) within its own encrypted credential store; ensure the underlying n8n instance itself is properly secured given the sensitivity of what it holds.
Backing Up n8n Workflows
See How to Set Up Automated VPS Backups — back up both the workflow definitions and any persistent database, ensuring you can recover your automation configuration in case of an issue.
Common Errors
Workflow execution fails at the LLM node with a timeout — verify your LLM endpoint (whether third-party API or self-hosted) responds within n8n's configured timeout window; self-hosted LLM inference (particularly on CPU) can sometimes exceed default timeout assumptions, requiring adjustment.
Continue Reading
- How to Install Ollama and Run Local LLMs on a VPS
- How to Install LocalAI as an OpenAI-Compatible API Alternative
- How to Build a Simple AI Chatbot with LangChain on a VPS
Browse more articles in AI & Machine Learning on a VPS.
