Short answer: Zapier has the widest integration library but no EU data residency. Make bills per module action and works out cheaply on long chains. n8n bills per workflow run, keeps cloud data in Frankfurt and can be self-hosted. For privacy-sensitive processes, n8n is usually the obvious pick.
The billing unit matters more than the list price
At first glance the three big automation platforms cost roughly the same. The difference does not come from the price tag, it comes from what gets counted. That is exactly where most of the calculations we see in first meetings fall apart.
n8n counts workflow executions. One run is one run, whether it contains three steps or thirty. The pricing page says so verbatim: an execution is a single run of your entire workflow, regardless of how many steps it has or how much data it processes.
Make switched its billing to credits. Each module action in a scenario normally consumes one credit, some AI-backed features consume more. Routers and error handlers are excluded. Build long chains and you pay more often.
Zapier counts tasks. A task is used when a Zap successfully moves data or completes an action. The trigger itself does not count, and neither do filters, formatters, paths or delays. AI steps carry multipliers of one to five.
For a business that means: a process with many small steps is cheapest on n8n, while a process with very many runs of only two steps each may work out better on Make or Zapier.
The comparison at a glance
| Criterion | n8n | Make | Zapier |
|---|---|---|---|
| Billing unit | workflow execution | credit per module action | task per action |
| Free entry point | self-hosted Community Edition | 1,000 credits/month | 100 tasks/month |
| Documented entry plan | Starter EUR 20/month, billed annually | Core from USD 12/month | Professional from USD 19.99/month, billed annually |
| Included volume | 2,500 executions | 10,000 credits | 750 tasks |
| Next tier | Pro EUR 50/month, 10,000 executions | Pro from USD 21/month | Team from USD 69/month |
| Integrations (vendor figure) | 2,053 | 3,000+ | 9,000+ |
| Cloud data location | EU, Frankfurt | AWS EU or North America | AWS USA, no EU option |
| Self-hosting | yes | no | no |
| Licence model | Sustainable Use License (fair-code) | proprietary SaaS | proprietary SaaS |
Worked example: 500 leads a month
Take a typical process. A website form fires the trigger, then six actions follow: create record, check for duplicates, write to CRM, send confirmation email, post to Slack for sales, append a row to a sheet. Volume: 500 enquiries per month.
- n8n: 500 executions. That fits comfortably into the Starter plan with 2,500 executions. Cost: EUR 20 per month, with roughly 2,000 executions left over for other workflows.
- Make: 500 times 6 actions equals 3,000 credits. That fits into Core with 10,000 credits. Cost: from USD 12 per month.
- Zapier: the trigger does not count, so 500 times 5 equals 2,500 tasks. The documented Professional entry point covers 750 tasks and is therefore not enough. Zapier scales task volume with a slider and only reveals the price for the 2,500 tier once you select it, so we deliberately quote no figure here.
The lesson is not that n8n is always cheaper. It is that you should model your real process before you book a plan. A second workflow with 40 steps and 50 runs a month flips the result immediately.
Data protection: where many projects are actually decided
Zapier is refreshingly blunt here. Asked in its own data privacy documentation whether data can be stored in the EU only, the answer is that Zapier does not support this option, and that data sits on AWS servers in the United States. Transfers rely on the EU-US Data Privacy Framework and standard contractual clauses; DP-Dock GmbH in Hamburg is named as the Article 27 GDPR representative. None of that is automatically unlawful, but it is a decision that has to be documented and justified.
n8n states for its hosted plans that data is stored within the EU on servers located in Frankfurt, Germany. With self-hosting you decide. Make runs its infrastructure on AWS in the EU or North America according to its own comparison table and documents EU zones in its API structure.
If personal data about applicants, patients or clients passes through an automation, this is the point where tool selection stops being about convenience. We describe how to approach that systematically in GDPR-compliant AI tools.
Self-hosting n8n: what is allowed and what is not
n8n is explicitly not an open source project. The company writes in its own documentation that, according to the OSI, open source licences cannot include limitations on use, and that it therefore does not call itself open source. What applies instead is the Sustainable Use License.
In practice, for a mid-sized company: you may run n8n on your own server, use it for internal business purposes, build your own nodes, and even charge clients for building workflows as a service provider. What is not allowed is white-labelling n8n or hosting it and charging people for access. Enterprise features require a separate licence anyway; the corresponding source files are carved out of the Sustainable Use License.
Self-hosting is not free, incidentally, it is just paid differently. A small server, backups, updates and somebody who feels responsible realistically cost more than EUR 20 a month. For companies without their own IT we offer this as a managed service.
Which tool for which company?
Zapier
Sensible when you need to connect a lot of niche applications and data protection is not critical, for example purely internal notifications without personal data. The breadth of integrations is unmatched and setup is the easiest of the three.
Make
The middle ground. Visual editor, solid error handling, EU infrastructure available. A good fit when you need more complex branching but do not want to run a server.
n8n
The choice when data sovereignty matters, when workflows have many steps, or when you want to embed your own scripts and AI components. The entry curve is more technical, the cost curve flatter as complexity grows.
Which process is worth automating first is a question best settled before the tool question. Our article on five processes small teams should automate right away gives you an order of play.
Frequently asked questions
Is n8n really free?
The self-hosted Community Edition is available at no cost. Costs come from the server, maintenance and your own time. Cloud plans start at EUR 20 per month billed annually (as of September 2026).
Can we migrate from Zapier to n8n later?
Technically yes, but there is no automatic importer. Expect to rebuild every workflow once. For five to ten automations that is one to two working days, and it is usually a good opportunity to clean house.
What happens if we exceed our quota?
Vendors handle this differently: throttling, add-on packs, or automatic upgrades. Make charges 25 percent more for extra credits than for plan credits according to its own release notes. Check this before you book, otherwise your first traffic spike gets expensive.
Do we need a data processing agreement for these tools?
As soon as personal data is processed, as a rule yes. All three vendors provide the relevant contract documents. Check the sub-processors and the data location as well.
Is automation worth it with ten employees?
Usually yes, but rarely where you look first. The benefit lies less in hours saved than in errors avoided and faster response times. You will find concrete examples in our overview of AI automation use cases for SMEs.
Conclusion
The tool question is the second question. The first is which process actually hurts and how often it runs. Once you know that, the platform almost picks itself: Zapier for breadth, Make for visual comfort, n8n for data sovereignty and depth. If you would rather not decide alone, take a look at our automation services or start with the AI check. All prices in this article come from the vendor pages n8n, Make and Zapier and were retrieved on 17 August 2026; plans change frequently.
