Cheapest option as execution volume grows
Steps per run and server cost held at your current inputs โ only monthly executions change per row.
| Executions/mo | n8n Cloud | n8n Self-Hosted | Zapier | Make | Cheapest |
|---|
Four billing shapes for the same workflow
n8n Cloud doesn't meter individual steps at all โ it counts executions, where one full workflow run (however many nodes it touches) is one unit, and bills a flat tiered plan price for a block of included executions per month (Starter ~2,500, Pro ~10,000, Business ~40,000, all on annual billing). There's no per-unit overage rate; once you clear a tier's allowance you upgrade to the next one, so the bill is a step function of volume rather than a smooth line. Self-hosting the open-source Community edition removes the meter entirely: the software is free with unlimited executions and workflows, so your only cost is the server it runs on โ a small VPS handles a surprising amount of volume before you need n8n's queue mode with separate worker processes. Zapier and Make both meter at the step/module level rather than the workflow level, which is why a long multi-step n8n workflow can look dramatically cheaper on n8n than on either โ n8n charges the same one execution whether the workflow has 3 nodes or 30, while Zapier and Make charge more for every additional step.
The practical takeaway: at low volume with simple workflows, n8n Cloud's Starter tier and Make's cheap per-operation rate are both hard to beat. As either execution volume or steps-per-workflow climbs, self-hosted n8n's flat server cost pulls further ahead of every metered competitor, because your bill stops scaling with usage at all. Compare this against the Zapier vs Make Cost Calculator for a deeper breakdown of exactly how those two platforms count units, or the Self-Host vs API Calculator for the same build-vs-buy trade-off applied to LLM inference.
How this calculator works
The n8n Cost Calculator compares four workflow-automation pricing shapes at your monthly execution volume: n8n Cloud (flat tiered plan โ the calculator picks the cheapest tier whose included executions cover your volume, or flags Enterprise once you exceed the Business tier), n8n self-hosted (your entered server cost, flat regardless of volume since the Community edition has no execution limit), Zapier (base plan + max(0, tasks โ included) ร overage rate, where tasks = executions ร steps per run) and Make (the same formula using operations = executions ร (steps per run + 1), since Make also counts the trigger). All four totals are shown side by side, then a second table isolates every option across a range of execution volumes at your current steps-per-run and server cost, so you can see exactly where self-hosting starts winning.
These are representative figures based on n8n's published Cloud pricing (Starter/Pro/Business tiers, annual billing) and the same Zapier/Make reference rates used across our automation calculators, not live vendor quotes โ n8n periodically adjusts tier boundaries and its Enterprise pricing is never published (buyer reports put it around $2,000-3,000/mo). Every field is editable: enter your own plan prices, included volumes, step counts and server cost for a number that matches your actual account.
Frequently asked questions
Is n8n cheaper than Zapier and Make?
Almost always, and the gap widens fast with volume. n8n Cloud's Starter plan (about $20/mo for 2,500 executions billed annually) already undercuts Zapier's per-task pricing at the same workflow volume, and self-hosting the free Community edition removes the per-execution meter entirely โ you only pay for the server, typically $6-60/mo regardless of how many workflows run on it. Make.com is the closest competitor on price at low-to-mid volume because its per-operation rate is very low, but once you're running tens of thousands of executions a month, self-hosted n8n on a small VPS is usually the cheapest option by a wide margin.
What is the difference between n8n Cloud and self-hosting n8n?
n8n Cloud is a managed service: n8n runs the servers, handles updates and scaling, and bills you a flat monthly fee per tier based on how many workflow executions you use (Starter 2,500/mo, Pro 10,000/mo, Business roughly 40,000/mo). Self-hosting means you run the open-source n8n Community edition yourself on your own server โ the software is completely free with unlimited executions, unlimited workflows and no per-run meter, but you're responsible for hosting, updates, backups and uptime, and very high-volume workloads may need a bigger server or n8n's queue mode with multiple workers to keep up.
What counts as one n8n execution?
One execution is one complete run of a workflow, regardless of how many nodes or steps it contains โ a workflow with 3 nodes and a workflow with 30 nodes both count as exactly one execution when they run once. That's a fundamentally different unit than Zapier's per-task billing (one unit per action step) or Make's per-operation billing (one unit per module), which is why a long, multi-step workflow can look far cheaper on n8n's execution-based pricing than on either competitor.
Are these exact n8n, Zapier and Make prices?
The defaults reflect n8n's published Cloud pricing as of mid-2026 (Starter/Pro/Business tiers, annual billing) and the same representative Zapier/Make figures used across our automation pricing tools, not a live price feed โ all three vendors revise plans periodically and n8n's Enterprise tier (unlimited executions, SSO, custom support) isn't published at all, with buyer reports putting it around $2,000-3,000/mo. Every field is editable: plug in your own plan price, included volume and server cost for an exact number.