← All comparisons Automation

Make vs n8n

Per-operation credits vs whole-run executions, and who should actually self-host.

The billing models are the real difference, not the feature lists. Make bills per module operation, so a long scenario spends credits at every step. n8n bills per execution, meaning one workflow run costs one execution no matter how many steps it contains, and its community edition is free to self-host. Make is the easier on-ramp for non-technical teams; n8n rewards technical teams running heavy volume.

At a glance

Maken8n
Source modelClosed source SaaSSource available on GitHub, fair-code license
HostingMake's managed cloudn8n Cloud or your own server
Billing unitCredits, consumed per module operationExecutions, one per workflow run, unlimited steps
Free option1,000 credits per month, 2 active scenariosCommunity edition, self-hosted, no execution cap
Entry paid planCore, $12 per month billed annually, 10,000 creditsStarter, 20 euros per month billed annually, 2,500 executions
Next tiersPro $21, Teams $38, both at 10,000 creditsPro, 50 euros per month, 10,000 executions
Scheduling and limits15-minute minimum interval on Free, 1 minute on paidCaps are executions and concurrency, 5 concurrent on Starter, 20 on Pro
Custom codeLow-code first, limited scriptingJavaScript and Python code nodes built in
Integrations3,000+ standard appsFewer prebuilt nodes, plus HTTP and code for the rest
Best forNon-technical teams that want managed simplicityTechnical teams, high volume, data control

Pricing

Make's free plan includes 1,000 credits a month, two active scenarios, and a 15-minute minimum schedule interval. Paid plans at the 10,000-credit level run $12 a month for Core, $21 for Pro, and $38 for Teams, billed annually, with a 1-minute interval and unlimited active scenarios. Enterprise is custom. One thing to watch: Make moved from counting operations to credits in August 2025, and AI features can consume credits at different rates than ordinary module runs, so audit your usage after turning anything AI-flavored on.

n8n Cloud starts at 20 euros a month billed annually for 2,500 executions, with Pro at 50 euros for 10,000 executions and Business at 667 euros for around 40,000 executions plus SSO and a self-hosted option. Every plan allows unlimited steps per execution; the practical caps are execution volume and concurrency. The self-hosted community edition is free with no execution cap, but note the license: it is fair-code under a sustainable use license, not OSI open source. Internal business use is fine. Reselling hosted n8n as a service is not, which matters if you are an agency.

Run the math on your own workflows before choosing. A 20-step workflow that fires 1,000 times a month consumes roughly 20,000 Make credits, twice the entry plan's allowance, but only 1,000 n8n executions, comfortably inside the cheapest cloud tier. Reverse it for short scenarios at low volume and Make's $12 entry, or even its free tier, wins.

When to pick Make

Pick Make when nobody on the team wants to touch a server and the automations are modest: a few hundred runs a month, mostly short scenarios connecting common SaaS tools. The visual builder is polished, the 3,000+ app catalog covers most small-business stacks, and there is nothing to patch or back up.

It is also the safer handoff. If you build automations for a client who will maintain them without you, a managed platform with a friendly interface fails more gracefully than a self-hosted server nobody owns.

When to pick n8n

Pick n8n when volume is high, workflows are long, or data has to stay under your control. The per-execution billing gets cheaper relative to Make as workflows grow more steps, and self-hosting means client or customer data never leaves infrastructure you control, which some regulated businesses require before they will sign off.

It is also the better fit for technical teams. JavaScript and Python code nodes are built in, the HTTP node covers any API without waiting for an official integration, and the native AI agent nodes make it a reasonable base for LLM workflows. The traction is real; n8n's own pricing page lists over 200,000 GitHub stars.

Just be honest about capacity. If nobody on the team can maintain a server, use n8n Cloud rather than self-hosting, and price it against Make on your actual workload.

What most comparisons miss

The billing unit dominates total cost, and most comparisons skip it. Comparing $12 to 20 euros tells you almost nothing until you multiply your steps by your run counts under each model. Do that arithmetic on your three heaviest workflows first.

Fair-code is not open source. n8n's sustainable use license permits free internal use but restricts offering n8n itself as a hosted service to others. If you plan to embed it in something you sell, read the license before you build.

Self-hosting is not free either. The server is cheap; the real cost is that you become the ops team for updates, backups, and security. An unpatched automation server holding API keys to your whole stack is a liability, not a savings.

And workflows do not port. There is no clean migration between the two platforms, so whichever you pick, you are likely keeping it for years. Choose based on where your volume and team will be in two years, not where they are this month.

My verdict

For a non-technical business that wants automations running this week, I recommend Make. For technical teams, heavy volume, long workflows, or data that has to stay put, I recommend n8n, and it is what I run myself. If you are torn, price your three heaviest workflows under both billing models. The answer usually falls out of the arithmetic, not the feature lists.