n8n Pricing Comparison
n8n's sticker prices are the easy part. Here's the full-cost math.
Why n8n pricing is structurally cheap
n8n charges per workflow execution — one run of a whole workflow, however many steps it contains. Zapier and Make charge per task/operation: every step, every time. A 20-step order-processing workflow that runs 1,000 times a month is 1,000 executions on n8n and 20,000 billable tasks elsewhere. At any real complexity, n8n wins the unit math by an order of magnitude. (Full three-way comparison.)
The n8n price list (2026 ballpark)
| Option | Price shape | Fine print |
|---|---|---|
| n8n Cloud Starter/Pro | ~€20–50+/mo by execution tier | Managed infra; limited custom nodes |
| Self-hosted Community | €0 license + ~€10–50/mo server | Plus maintenance hours — the real line item |
| n8n Enterprise | Custom (four figures/yr up) | SSO, RBAC, environments, air-gapped (when it's worth it) |
| Managed/operated n8n | Fixed monthly fee | Infra and workflow operations included |
Exact tiers move; check n8n's pricing page for current numbers. The shapes above are what matter for the decision.
The line nobody budgets: maintenance
Self-hosted n8n's license price of zero is real, and so are the hours: updates, backups, TLS renewals, Postgres care, debugging the workflow that stopped Friday night. At any loaded hourly cost, a few hours a month exceeds the Cloud subscription you avoided — we did this math in detail. Self-hosting is a control decision, not a savings plan (the honest trade-off).
What managed n8n costs and covers
Oido's managed n8n is a fixed monthly fee scoped by instance size, workflows under management, and support level — deliberately not per execution, so growth doesn't inflate the bill. It covers hosting (ours or your infrastructure), monitoring of executions rather than just uptime, fixes, and engineering time for changes. Compared honestly: it costs more than a Cloud subscription and less than the fraction of an engineer's salary that DIY quietly consumes — that's the whole trade.
How to compare for your case
- Count your planned workflows × runs/month × steps — that exposes the per-task vs per-execution gap.
- Price the maintenance hours honestly (who, how many, at what loaded cost).
- Add what a silent failure costs you — an unprocessed order queue for a weekend is a pricing input too.
Want the numbers run on your actual processes? Free consultation — also see how much AI automation costs overall.