Huginn vs Node-RED

Huginn

• Developers who want their automation data kept on a server they control
• Watching websites, feeds and search terms and getting alerts by email or SMS
• People comfortable with Docker or a Ruby on Rails deployment

Node-RED

• Home automation and device projects on hardware such as a Raspberry Pi
• Industrial teams collecting and transforming operational data
• Developers who are comfortable running a Node.js service themselves

Huginn

• There is no hosted version: you install, update and back up Huginn yourself, with MySQL or PostgreSQL alongside it.
• Huginn does not run on Windows; the docs suggest a virtual machine. The manual install guide supports Ubuntu and Debian.
• The requirements page recommends 2 GB of RAM and a dual-core server.
• Agents are configured in the web interface with options and Liquid templates, and writing a new agent means writing Ruby.

Node-RED

• Node-RED is software you install and run yourself; the project offers no hosted service of its own, though third parties such as FlowFuse do.
• The machine that runs it needs a supported version of Node.js, or Docker.
• The project calls it low-code: custom logic is written as JavaScript in function nodes.
• Running on Android through Termux is described in the docs as a bit experimental.

Huginn

Agents and events

Each agent does one job, such as checking a site or sending a message, and emits events that other agents consume. Chaining them forms a directed graph.

Website monitoring

Agents can scrape websites and email you when a page changes, or watch for travel and shopping deals. The wiki covers browser emulation for pages that need JavaScript.

Digests and spike alerts

Huginn can send a digest email at set times of day, and count high-frequency events so that an SMS goes out within moments of a spike.

Webhooks and custom JavaScript

Agents send and receive webhooks and can run custom JavaScript functions, so services without a built-in agent can still be wired in.

Agent gems

New agents can be written as external Ruby gems and loaded through the ADDITIONAL_GEMS environment variable, without changing the core repository.

Several ways to deploy

There is an official Docker image, a manual installation guide for Ubuntu and Debian, OpenShift templates and a Heroku deploy button, which needs a paid Heroku plan.

Node-RED

Browser-based flow editor

Flows are built by wiring together nodes from a palette in the browser, then deployed to the runtime in a single click.

Function nodes

JavaScript functions can be written inside the editor in a rich text editor, and a built-in library saves functions, templates and flows for reuse.

Community nodes and flow library

The palette can be extended with new nodes, and the project says the community has shared more than 5,000 nodes and flows. An online flow library hosts them.

Runs from edge devices to the cloud

The Node.js runtime is light enough for low-cost hardware such as the Raspberry Pi and BeagleBone, and guides also cover Docker, Windows, Amazon Web Services and Microsoft Azure.

Flows stored as JSON

Flows are saved as JSON, which can be imported and exported to share with others. A Projects feature, which needs git, puts flows under version control.

APIs and embedding

An Admin HTTP API administers the runtime remotely, and module APIs let developers embed Node-RED in an existing Node.js application.

Huginn

• Huginn$0

  The whole project under the MIT License: the web interface, the built-in agents and support for agents added as Ruby gems.

Node-RED

• Node-RED$0

  The whole project under the Apache 2.0 license: the runtime, the browser-based editor and access to the community's nodes and flows.