Blog
Tutorial10 min

How to Use n8n: Automate Anything Without Paying Per Run

Practical guide to n8n in 2026: installation, building your first workflow, connecting apps, and 10 automation templates. The open-source Zapier alternative.

May 25, 2026TheAISelect

n8n is the automation platform taking users from Zapier and Make for one simple reason: it's open source, you can self-host it, and you don't pay per execution. This guide takes you from zero to a working workflow.

Why n8n Instead of Zapier or Make?

n8n has three advantages that make it especially compelling in 2026:

  1. Open source and self-hosted: you control your data, no execution limits
  2. Code when you need it: every node can run JavaScript or Python for complex logic
  3. Native AI integration: built-in nodes for OpenAI, Anthropic, Hugging Face, and local models

The tradeoff: the learning curve is steeper than Zapier. For simple automations, Zapier is faster to set up. For complex workflows or high volume, n8n wins easily.

Getting Started: Two Options

Option A: n8n Cloud (easiest)

  1. Sign up at n8n.io — free plan includes 20 executions/month
  2. Access the interface from your browser
  3. Start building workflows immediately

Option B: Self-hosted with Docker (recommended for serious use)

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -v ~/.n8n:/home/node/.n8n \
  docker.n8n.io/n8nio/n8n

Access at http://localhost:5678. Cost: $0. Executions: unlimited.

How n8n Workflows Work

A workflow is a graph of nodes where each node:

  • Receives data from the previous node
  • Executes an operation (send an email, call an API, filter data)
  • Passes results to the next node

The interface is drag-and-drop. Drag nodes from the left panel, connect them, and configure each one.

The Most Important Nodes

NodeWhat it does
WebhookReceive data from any external app
Schedule TriggerRun workflow on a fixed schedule
HTTP RequestCall any API
IFBranch based on a condition
SetModify or create data fields
CodeCustom JavaScript or Python
OpenAI / AnthropicCall AI models

Your First Workflow in 10 Minutes

Goal: send a daily summary of industry news to your email every morning.

  1. Trigger: Schedule → every day at 8:00 AM
  2. HTTP Request: fetch an RSS feed from an industry news source
  3. Code: extract the 5 most recent headlines (3 lines of JavaScript)
  4. OpenAI: summarize and prioritize the headlines with a prompt
  5. Gmail / SMTP: send the summary to your email

Total: 5 nodes, 15 minutes of setup, and you never have to manually check news again.

10 Automations Worth Building

1. Brand mention monitoring

Webhook → search Twitter/X and Reddit → filter negative mentions → notify in Slack

2. Automated lead pipeline

Web form → enrich data with an API → add to CRM → send personalized welcome email

3. Weekly analytics digest

Schedule (Friday 5 PM) → pull GA4 data → generate summary with AI → email to team

4. Support email classifier

Gmail → read new email → OpenAI classifies urgency and category → assign in Jira/Linear

5. Auto-post to social media

Blog RSS → when new post published → generate copy variations with AI → post to Twitter, LinkedIn

6. Competitor price monitor

Daily schedule → scrape competitor price (HTTP Request) → IF price changed → alert in Slack

7. Automatic meeting transcription

Webhook from Zoom/Google Meet → download recording → Whisper transcribes → Claude summarizes + extracts action items → save to Notion

8. Invoice generator

Trigger on new client in CRM → generate PDF invoice → send by email → mark as sent in CRM

9. Automated data backup

Weekly schedule → export data from all apps → compress → upload to Google Drive or S3

10. Knowledge base chatbot

Webhook receives question → search Pinecone (vector DB) → OpenAI answers with context → return response

Connecting n8n to AI

n8n has native nodes for OpenAI, Anthropic, and Hugging Face. The basic flow:

  1. Add an OpenAI or Anthropic node
  2. Select the model (GPT-4o, Claude Sonnet, etc.)
  3. Configure the prompt with workflow variables: Analyze this text: {{ $json.content }}
  4. Process the result in the next node

For RAG (semantic search), n8n has native Pinecone and Qdrant nodes that work out of the box.

n8n vs Zapier vs Make

n8n wins when: high volume, complex logic, data privacy requirements, cost at scale.
Zapier wins when: ease of use, non-technical team, fast setup for simple tasks.
Make is in between: more powerful than Zapier, easier than n8n.

See the n8n vs Zapier comparison and the full n8n review.

Tags#tutorial#n8n#automation#no-code