Lesson 8 · 2. Getting Started · Last updated · By Dheeraj Sharma

n8n Node Types Explained (2025) | How to Build Workflow with Triggers, Apps, Core & Actions

Learn n8n Node Types Explained (2025) | How to Build Workflow with Triggers, Apps, Core & Actions in this comprehensive n8n tutorial. Step-by-step guide with video, code examples, and expert tips from the n8n Zero to Hero course.

Table of Contents

Overview

This lesson is part of the n8n AI Automation - Zero to Hero course, Section: 2. Getting Started.

Watch the video above for the full tutorial, or read the written guide below.

What are the four node types in n8n?

n8n classifies every workflow component into one of four node types. Trigger nodes start the automation when a specific event fires, such as a form submission or a new Slack message. App nodes (also called action nodes) add, remove, or send data to external systems. Core nodes handle logic, scheduling, and generic API calls without connecting to an external service. Cluster nodes group multiple nodes into one unit, like an AI agent, and are covered in the AI-integration section of the course.

How to build a workflow using trigger, app, and core nodes in n8n

Trigger nodes set the starting condition for every production workflow. The simplest option is the Manual Trigger, which fires when you click "Test Workflow." Real automations use event-based triggers instead: a new row added to Google Sheets, a new email in your inbox, or a scheduled time such as every morning at 8 a.m. Every workflow you deploy in n8n must include exactly one trigger node, because that node decides when the automation runs.

App and core nodes sit in the middle of the workflow, processing data between the trigger and the final output. To add one, click the "+" icon on the canvas and search for the service or function you need. Selecting Google Sheets opens the node configuration panel, where you pick the operation (such as "Get Rows"), connect OAuth credentials, and choose the target document and sheet. The panel shows incoming data on the left, configuration settings in the center, and output on the right.

Action nodes complete the workflow by sending results to an external destination, such as updating a spreadsheet, posting a Slack notification, or writing to a database. Once a node runs successfully, you can click "Pin Data" to freeze its output for the rest of your build session, so you do not have to re-execute that node on every test run. n8n charges per workflow execution rather than per node, so pinning data saves build time without adding cost.

Key Takeaways

  • Trigger nodes listen for a specific event (form submission, scheduled time, new data row) and every deployed n8n workflow must have exactly one.
  • App nodes connect to named external services like Google Sheets or Slack to read, write, or trigger actions in those systems.
  • Core nodes (Filter, Set, Code) transform and route data using logic or custom JavaScript/Python without authenticating to a specific third-party service.
  • Cluster nodes, such as the AI Agent node, bundle multiple nodes into a single unit and are covered separately in the AI-integration portion of the course.
  • Pin Data locks a node's output during development, letting you test downstream steps repeatedly without re-fetching live data on every run.

Next Steps

Continue your n8n journey with the full n8n AI Automation - Zero to Hero course.

Frequently Asked Questions

Trigger nodes are the only nodes that decide when a workflow runs, not what it does. While app and core nodes process data that is already flowing, a trigger node waits for a specific event, such as a new email, a form submission, or a scheduled time, then starts the entire execution chain. Every deployed n8n workflow must begin with exactly one trigger node.

Core nodes perform logic, scheduling, or generic API calls entirely within the workflow without authenticating to a specific named external service. App nodes, by contrast, call a named third-party service such as Google Sheets or Slack. Examples of core nodes include the Filter node (conditional checks), the Set node (adding or modifying fields), and the Code node, which accepts both JavaScript and Python.

The Pin Data feature freezes a node's output so n8n reuses those cached results for every downstream test run instead of re-executing the node. In the course demo, pinning the Google Sheets 'Get Rows' node locked 100 sample rows in place across the entire build session. A blue icon appears on the node to confirm its output is pinned.

Cluster nodes group multiple individual nodes into a single combined unit to accomplish a complex task, with the AI Agent node being the primary example given in the course. The instructor classifies cluster nodes as the fourth node type but explicitly defers them to the AI-integration section, so they are not configured in the introductory node-types lesson.

Dheeraj Sharma

Dheeraj Sharma

AI Systems Builder
Creator of the n8n Zero to Hero course (42 lessons, 31+ hours). I help solopreneurs build AI systems that grow revenue without growing workload.

Get the n8n Mastery Bundle

All workflows, cheat sheets, and premium resources from the entire course in one package.

Get Premium Resources