n8n
Copying information between apps by hand wastes hours. n8n lets you connect your apps and AI on a visual canvas, so a routine task runs on its own, with your approval where you want it.
How you use it: Use the hosted version in your browser or install it with Docker, then start from a ready-made template.
In plain English
The README describes n8n as a fair-code platform for AI agents and workflows. You build on a visual canvas, and you can drop into JavaScript or Python when a step needs code. It says you can run it on your own server or in the cloud.
The README says 1500+ integrations and 9,000+ workflow templates, and it says you can connect OpenAI, Anthropic, Google, or open-source models and switch providers. The name is pronounced n-eight-n. The founder note in the README says it comes from nodemation.
The license is not a standard permissive open-source license. The README says n8n is fair-code under the Sustainable Use License, with a separate Enterprise License. You can see the source and self-host it. Read the license page it links before you use it in a business.
Practical ways to use it
Example 1
Connect an AI step to tools you already use
The README points at a large integration list and a template gallery, so a workflow can take an input, ask a model to draft a reply, and file the result.
Example 2
Keep a human in the loop
The capabilities list includes multi-step AI workflows with human approvals, which fits a draft-then-review process.
Example 3
Avoid locking the workflow to one model vendor
The README says you can connect several model providers, including open-source models, and switch without changing the whole design.
Example 4
Host it yourself when the data is sensitive
Self-hosting is a headline feature. The cloud is the alternative when you do not want a server.
Who it is for
Non-coder. The product is a visual builder, which is the non-coder path. The quick start on your own machine uses Docker. The cloud login is the path that skips Docker.
Difficulty: Takes some setup. The cloud account avoids the install. Self-hosting needs Docker, then the editor is the friendly part.
Getting started
- 1
Decide cloud or self-host
The README links the hosted editor and shows a Docker install for your own machine. Pricing for the cloud is not printed in the README, so check n8n.io before you subscribe.
- 2
Self-host with Docker
The README says the install script needs Docker, then shows a curl install from get.n8n.io. A manual docker run is also printed, publishing port 5678.
- 3
Open the editor
The README says the editor is at http://localhost:5678 after the Docker run.
- 4
Start from a template and read the license
Browse the workflow templates linked in the README. Open the Sustainable Use License before you build a product on top of n8n.
What you need first
- Accounts
- The cloud path needs an n8n account. Self-hosting needs Docker and whatever accounts the integrations in your workflow use, such as a model API key.
- Hardware
- Docker on the self-host path. The README does not list CPU or RAM minimums in the quick start.
- Cost
- Self-hosting costs you the computer plus any model or app fees in the workflow. The cloud has its own plan. The README does not paste prices.
Cautions
- Fair-code is not the same as MIT. The Sustainable Use License and the Enterprise License are both named. Do not assume you can resell the platform.
- A workflow can send messages and change records in other apps. Test with harmless data.
- The install script runs a remote script. Docker is a prerequisite either way. Use the manual docker run if you want to see each flag.
Questions people ask
What does n8n do?
It lets you build AI workflows and automations on a visual canvas, then run them on your own server or in n8n's cloud.
Who is n8n for?
People who want a visual builder and may not write code every day. Self-hosting still asks you to use Docker, or you can use the cloud editor.
What do I need first?
Docker for the self-host quick start, or an account at the cloud link. Model providers and other apps need their own keys when a workflow calls them.
What should I watch out for?
Read the Sustainable Use License before business use. Workflows can take real actions, so test them on harmless data first.