How to learn n8n in 2026

What is n8n?

n8n (said “n-eight-n”) lets you build automations on a visual canvas. Each workflow starts with a trigger — a form is submitted, an email arrives, it is 9am on Monday — and then runs a chain of nodes that read, change and send data between your apps.

It connects to hundreds of apps out of the box, and to almost anything else through its HTTP Request node. What made it popular with AI builders is its AI nodes: you can drop a language model, an AI agent with tools, or a search over your own documents straight into a workflow.

Do you need to know how to code?

No. Most workflows are built by dragging nodes and filling in fields. Two things help, and both are learnable in an afternoon:

• Reading JSON — the shape data takes as it moves between nodes.
• Simple expressions — small snippets in {{ }} that pull a value from an earlier step, like a customer’s name.

There is a Code node for JavaScript or Python when you need it, but plenty of people use n8n for years without opening it.

n8n Cloud or self-hosted?

For learning, start on the cloud trial or run it locally on your laptop. Decide about hosting once you have a workflow worth keeping.

A 4-week plan to learn n8n

1. Week 1 — The basicsTriggers, nodes, credentials and how data flows as JSON. Build: a form that saves each reply to Google Sheets and sends you an email.
2. Week 2 — Data and logicIF and Switch to branch, Merge to combine, loops over many items, and expressions to reshape data. Build: a lead tracker that sorts new leads and alerts the right person.
3. Week 3 — AI agentsAdd a chat model, then an AI Agent node with memory and tools (search, a sheet, an API). Ask the model for structured JSON so the next node never breaks. Build: a support agent that answers common questions and hands the rest to a human.
4. Week 4 — Your documents, and going liveLoad your PDFs into a vector store and let the agent answer from them, with sources. Then add error workflows, a human approval step before anything reaches a customer, and run it on a schedule. Build: a document chatbot your team can use.

Five beginner projects worth building

• Lead pipeline — form → AI scores the lead → message to your team → row in your CRM.
• Meeting notes → action items — transcript in, decisions, owners and dates out.
• Weekly report — numbers pulled on a schedule, a short written summary emailed every Monday.
• Content pipeline — one long post turned into drafts for LinkedIn, a newsletter and short tips.
• Document chatbot — answers questions from your own PDFs and shows where each answer came from.

Mistakes that slow beginners down

• Starting with the impressive workflow instead of the boring, frequent one.
• Letting an AI node return free text — ask for JSON and check it.
• Sending anything to a customer without a human approval step.
• No error workflow, so a failed run goes unnoticed for days.
• Watching tutorials for weeks without shipping a single workflow.

Frequently asked questions