Node-RED

What is Node-RED?

Node-RED lets you build automation by dragging nodes onto a canvas and wiring them together, with a huge library of community nodes for MQTT, HTTP, home automation hardware and cloud services. It started inside IBM as a way to make event-driven IoT work visual, and it is still the quickest way to connect a sensor or webhook to an action without writing a full service.

Best for

Wiring sensors, webhooks and APIs together visually without writing much code

Why choose Node-RED

Node-RED is the most natural way to wire things together if you think visually and your inputs are devices. It was built for connecting hardware and APIs, and it shows: MQTT and HTTP nodes are first class, function nodes drop you into JavaScript when logic demands it, and the community's node library covers nearly every device and service you can name. For anything involving a sensor, a message broker or a stream of events, it is faster to build in than a general automation platform.

Replaces

• Zapier
• IFTTT
• n8n

Key features

• Visual flow editor that runs in the browser
• Large library of community-contributed nodes
• Built-in MQTT, HTTP and websocket support
• JavaScript function nodes for custom logic

What to watch out for

The visual metaphor stops scaling gracefully. A flow with a few dozen nodes is a pleasure; one with several hundred becomes a diagram you have to study before every change, and the built-in version control story for flows is weak. Debugging is by watching messages move, which is effective but manual. The palletised node library also means quality varies widely between contributors — a hastily written community node can be the least reliable part of your setup.

How to deploy

• Docker
• npm
• Linux packages

Getting started

Run it in Docker and put the flow file somewhere backed up, since every automation you build lives in that one file. Turn on the built-in security settings before exposing the editor, or anyone who reaches the port can rewrite your house. Install community nodes sparingly and check when they were last updated. Start by building one small flow that reads from a real device or API and logs the result, so you learn how messages carry their payload and topic before you write anything that acts on the world.

Typical setup

It typically runs in Docker on a home server or a small industrial box, wired to MQTT brokers, sensors, cameras, calendar APIs and chat services. Home users connect it to smart-home hardware; small businesses use it for device monitoring and alerting. Because the whole flow lives in one file, people back that file up carefully, and many keep a copy in version control to see what changed when something stops behaving.

Who should look elsewhere

Avoid it for large, long-lived workflows with many contributors. The visual model stops scaling around a few hundred nodes, and version control for flows is weak, which makes collaborative editing awkward. Teams building complex integrations with tests and code review should write code, not wire diagrams.

Project health

• GitHub stars: 23,699
• Last code push: 2026-09-18
• Open issues: 356
• Status: actively developed

Figures pulled from the GitHub API and refreshed periodically.