Getting Started with n8n

What n8n is

n8n is a fair-code workflow automation platform for building AI agents and automations. You design workflows on a visual canvas — nodes for triggers, apps, logic, and AI models — and extend any node with JavaScript or Python when the visual tools aren't enough. It ships with AI agent nodes, connects to major language model providers, and offers thousands of workflow templates.

What sets n8n apart from pure no-code tools is the combination: visual building for speed, real code for power, and the option to self-host the whole thing on your own infrastructure. Our no-code agent guide covers the general approach; this guide is the n8n-specific path. See how n8n compares to its closest rival in our n8n vs Lindy head-to-head.

Self-hosted vs n8n Cloud

You have two ways to run n8n, and the choice shapes both cost and responsibility:

• Self-hosted (Community Edition): free to license, unlimited executions, and your data stays on your server — as long as you also self-host the AI models you call, since calls to external model APIs send data to those providers. You're responsible for installation, updates, backups, and security. A good fit if you're technical or have data-residency requirements. See our open-source vs commercial comparison for the trade-offs.
• n8n Cloud: hosted by n8n — sign up and build in minutes, no server to maintain. Billed per workflow execution, which makes it the faster start for non-technical teams.

Get running in 5 steps

1. Pick your route. Technical and cost-sensitive? Self-host. Want the fastest start? n8n Cloud.
2. Install (self-hosted). With Docker installed, the official quick start is:

   docker volume create n8n_data
   docker run -it --rm --name n8n -p 5678:5678 -v n8n_data:/home/node/.n8n docker.n8n.io/n8nio/n8n

   Then open http://localhost:5678 in your browser and create your owner account.
3. Add your model credentials. Connect your AI provider (OpenAI, Anthropic, Google, or a local model) under Credentials — this powers the AI nodes. Model usage is billed by the provider, separately from n8n.
4. Build your first workflow. Start from a template: a scheduled trigger, one app action, and an AI node that summarizes or drafts. Run it manually first, then activate it.
5. Try an AI Agent node. Add the AI Agent node, give it a clear system prompt, connect two or three tools, and test with real inputs before letting it run on a schedule. Keep approval steps in for anything that sends messages or changes records.

Pricing at a glance

As of October 2026, n8n Cloud starts at €20/mo, billed annually (n8n.io/pricing, verified Oct 8, 2026) for 2,500 executions (Starter); the self-hosted Community Edition is free with unlimited executions. Higher tiers and add-on features — check the official pricing page for current plan details.

The key pricing insight: n8n bills per execution — one full workflow run, however many steps it contains — not per step. A 20-node workflow costs one execution. That's why complex automations stay cheap on n8n, and why a polling trigger that fires constantly is the classic budget trap. Our pricing models guide breaks down how this compares to task-based billing.

Good first workflows

• Email triage: classify incoming mail, draft replies, escalate the tricky ones.
• Lead enrichment: when a form submits, research the company and add notes to your CRM.
• Content repurposing: turn a transcript into a draft post and social summaries.
• Daily briefing: each morning, pull calendar, tasks, and key metrics into one summary.
• Database sync: keep two systems in agreement without manual copy-paste.

See our affiliate disclosure for how we fund independent reviews.