What is n8n? A Plain-English Explanation | 2V Automation
n8n is a workflow automation platform you can self-host, write real code in, and run as production infrastructure. Here's what it actually is and where it fits.
n8n is a workflow automation platform. You drag boxes onto a canvas, wire them together, and each box does one thing — fetch data from an API, transform it, call an AI model, write to a database, send a message. A connected chain of boxes is a workflow, and the platform runs that workflow whenever it’s triggered.
Pronounced “n-eight-n” (short for “nodemation” — node automation), it was founded in Berlin in 2019 and has become the tool of choice for teams who’ve outgrown Zapier and Make and want real control over their automation.
This post is the short answer. If you want the full version — the engine internals, production deployment, AI primitives, all of it — read our complete n8n automation guide.
What n8n is, in one sentence
A source-available, self-hostable, node-based workflow automation engine that you can program with real JavaScript and Python — not just visual drag-and-drop.
What it actually does
The mechanics are simple:
- A trigger fires. A webhook, a schedule, an email landing, a form submission, a database row inserted, a manual click.
- Items flow through nodes. Each node receives data, does one thing with it, and passes the result downstream. You can branch, merge, loop, filter, and transform along the way.
- The workflow finishes. Records get created, messages get sent, files get moved, dashboards get updated. The execution is logged with full input/output history for debugging.
The visible layer is a canvas where you connect nodes. The invisible layer is a Node.js runtime that executes the graph against your data.
What makes it different
There are dozens of workflow automation tools. n8n is interesting for four specific reasons:
1. You can self-host it
The Community Edition runs on your own infrastructure — a Docker container on any VM, a Kubernetes cluster, your own machine. You own the database, the credentials, the data in flight, the runtime. You’re not renting compute, you’re running software you control.
This matters when:
- You have data that can’t legally leave your infrastructure
- You’re going to run high enough volume that managed-SaaS pricing gets uncomfortable
- You want to write custom code that hooks into internal libraries
- You already have an ops team and one more service costs you nothing
2. You can write real code
Every n8n node accepts inline JavaScript expressions in any field. A dedicated Code node runs full JavaScript or Python on the items flowing through the workflow. When the pre-built nodes don’t do what you need, you write the missing logic right there on the canvas. No artificial ceiling.
Compare to Zapier, where Code by Zapier is sandboxed and severely limited, or Make, where custom code lives mostly on Enterprise plans.
3. The pricing model is per-execution, not per-task
A workflow that processes 200 records counts as one execution on n8n. On Zapier, it counts as 200 tasks. At any kind of scale, this difference is the difference between a $50 bill and a $2,000 bill.
The self-hosted Community Edition has no software fee at all — you pay for your VM, full stop.
4. Workflows are JSON you can put in Git
Every n8n workflow can be exported as JSON. You can diff it, code-review it, store it in source control, promote it between environments. This is the closest thing to “infrastructure as code” in the no-code automation world. It matters as soon as more than one person is building workflows.
Where n8n sits in the landscape
| Tool | Hosting | Code | Best for |
|---|---|---|---|
| Zapier | SaaS | Sandboxed | Fastest first zap, simple triggers |
| Make | SaaS | Limited | Visual logic for non-developers |
| n8n | SaaS or self-hosted | Full JS/Python | Production ops, AI workflows, scale |
| Power Automate | Microsoft cloud | Limited | Microsoft 365-heavy environments |
| Workato | SaaS | Limited | Enterprise integration with strong governance |
| Apache Airflow | Self-hosted | Python-first | Data pipelines, ETL jobs |
n8n’s sweet spot is the band between “simple zap” (where Zapier wins on speed-to-first-automation) and “Python service” (where engineers write code from scratch). It’s where most operations-heavy teams end up.
What n8n is used for
A non-exhaustive list of what we and our clients build on it:
- Syncing data between CRMs, marketing tools, billing systems, and internal databases
- Inbound webhook handling for SaaS integrations at moderate-to-high volume
- AI agents and assistants — Slack bots that answer questions from internal docs, support triagers that read tickets and draft replies, research agents that enrich leads before they hit a rep
- Document extraction pipelines — PDFs in, structured data out, into accounting or CRM
- Internal “glue” scripts that used to live in a Python repo nobody maintains
- Scheduled jobs — daily syncs, hourly reports, weekly cleanup
- Approval workflows that route through Slack or email
- Cross-tool dashboards that pull from APIs and refresh on a schedule
If you can describe a process as “when X happens, do Y, then Z, except when…” — that’s a workflow, and n8n is going to be able to build it.
The license situation, briefly
n8n’s source is available, but it’s not strictly open source under the OSI definition. Since late 2022 it has been licensed under the Sustainable Use License, which lets you use, modify, and self-host the code for any internal business purpose at no cost. The restriction: you can’t take the code and stand up a competing managed-n8n SaaS.
For 99% of use cases — running internal workflows for your business — this functions identically to open source. The distinction only matters if you were planning to compete with n8n the company.
What it costs
Two tracks:
- Self-hosted Community Edition: free. You pay for the VM you run it on (a $20/month box handles a lot of work). No execution limits, no feature gates beyond Enterprise extras like SSO and external secrets.
- n8n Cloud: paid, with multiple tiers based on execution volume and feature access. Current pricing is at n8n.io/pricing. The trade is zero infrastructure overhead in exchange for a monthly fee.
We’ve broken the cost decision down in detail in our n8n pricing 2025 breakdown.
When n8n is the wrong choice
To be fair to the alternatives:
- For your first-ever automation: Zapier is faster. If you need a Google Form to post to a Slack channel by lunch, use Zapier.
- For pure SaaS-to-SaaS plumbing in a non-technical team: Make’s canvas is genuinely beautiful and can be easier to reason about.
- For high-throughput data pipelines with strict orchestration semantics: Airflow or Temporal are designed for that. n8n can do a lot, but it’s not the right tool for ETL at scale or long-running workflows with exactly-once guarantees.
- For Microsoft-heavy environments where Power Automate is already in your tenant: the integration depth into Office 365 and Dynamics is hard to beat.
How to actually try n8n
The lowest-friction path:
- Sign up for a free n8n Cloud trial, or
- Run
docker run -it -p 5678:5678 n8nio/n8non your own machine
Either gets you into the editor in two minutes. Build something small that touches a real system you use — sync a Google Sheet row to a Slack message, push a Stripe webhook to a Notion database. Click around. You’ll know within an hour whether this is your tool.
Related reading
- n8n automation: the complete guide — the long-form pillar with engine internals, production deployment, AI primitives
- n8n vs Zapier: the complete comparison
- n8n vs Make vs Zapier: which automation tool wins?
- How to migrate from Zapier or Make to n8n
- Why your business needs its own n8n node
If you’re trying to figure out where automation will pay back first in your business — n8n or otherwise — the fastest answer is our Efficiency Scorecard. It’s 15 minutes, it’s free, and you keep the output regardless of whether we work together.