n8n

What n8n does

n8n is a workflow automation platform built around a visual node editor, with the important difference that you can run it on your own infrastructure. Workflows combine service integrations, conditional logic, loops and custom JavaScript or Python steps, and dedicated nodes build agents that call tools, query vector stores and chain model calls. Self-hosting means data stays inside your network and execution volume costs infrastructure rather than per-task fees, which changes the economics for high-volume automation. A managed cloud version exists for teams that prefer not to operate it. It suits technical teams, agencies building automation for clients and anyone with data residency requirements. The limitations are operational: self-hosting means you handle upgrades, backups and security, the node editor becomes hard to read as workflows grow, and the source-available licence restricts offering it as a competing hosted service.

How to use it

You can try the cloud version at n8n.io with a trial, or self-host with a container image, which is how most technical teams run it; a database and persistent storage are needed for anything beyond experimentation. Build your first workflow from a template, then open each node to see how data passes between steps, because understanding the item-based data model is the single thing that makes or breaks later work. Use the execution view to inspect the actual payload at each node rather than guessing at field names. For agent workflows, start with a single tool attached and add more only once the behaviour is predictable. Two operational points: pin credentials to environment variables rather than storing them in workflows you might export, and set up error workflows so failures notify you instead of failing silently. Cloud plans are priced by workflow executions, while self-hosting costs only your infrastructure. Version your workflows in a repository as exported JSON, so changes are reviewable and a broken edit can be restored quickly.

Best for

• self-hosted automation
• high-volume workflows
• AI agent pipelines
• data residency requirements
• technical teams

Strengths and limitations