CentriWebn8n vs Make is mostly a choice between two ways of paying and two ways of looking after it. Make charges credits for each action its modules perform and runs in its own cloud, in a US or EU data centre you pick once. n8n charges per execution, one full run of a workflow however many steps or records it handles, and can run on n8n’s cloud in Frankfurt or on your own server anywhere, including Australia.
A disclosure first: CentriWeb builds most client automations in n8n, so weigh our view with that in mind. Every vendor fact below comes from their own pages as at November 2026. Check their page for current prices. Our Zapier vs Make vs n8n comparison covers all three tools at a high level. This article goes further on the two that most often end up side by side.
n8n vs Make at a glance
| Make | n8n | |
|---|---|---|
| What it is | Visual automation platform, hosted by Make | Low-code workflow tool, hosted by n8n or self-hosted |
| How it charges | Credits: each module action, most use 1 | Executions: each full run of a workflow (cloud and paid plans) |
| Who maintains it | Someone comfortable with logic and data mapping | The same, plus code and server upkeep if self-hosted |
| Where data sits | US or EU data centre, chosen per organisation, fixed | Frankfurt on n8n Cloud, or wherever you host it |
| Who it suits | Branching flows at moderate volume, no server to run | Long or record-heavy flows, or data kept in a chosen country |
How Make counts credits
Make’s pricing page says each module action uses credits, and most actions use one. Some advanced features that use Make’s own AI provider can use more, and the page does not give figures for those. Actions that use credits include:
- reading data from an app or a webhook
- searching, creating, updating and deleting records
- transforming data with built-in tools
- aggregating and iterating
Router modules and error handler modules (Rollback, Break, Resume, Commit and Ignore) do not use credits. So branching and error handling are free to add, which encourages good habits.
Iterating is the part to watch. A module that runs once for each of 200 invoices performs 200 actions, and every module after the iterator repeats for each item. Run a test and check the scenario’s history to see what your real credit use is before you commit to a plan.
What happens when credits run out
The pricing page is clear about this, and it is one of Make’s better design choices:
- scenarios stop until you add credits
- incoming webhooks queue, up to your queue limit, and are processed once scenarios resume
- polling scenarios pick up records from their last successful run
- you can upgrade, buy extra credits in bundles of 1,000 or 10,000, or turn on automatic purchase of 10,000 credits on Core, Pro and Teams
- extra credits expire after one month on monthly plans, or at the end of the year on annual Pro and Teams plans
Credits also come with usage allowances. For every 10,000 credits a month, the page lists 5 GB of data transfer, 10 MB of data storage and up to 667 queued webhooks. A scenario that moves large files can hit the transfer allowance before the credit limit.
The free plan includes 1,000 credits a month. The page names Core, Pro, Teams and Enterprise plans and says annual billing is cheaper than monthly, but the version we opened did not show plan prices or credit amounts in a form we could read, so check the page directly for current figures.
How n8n counts executions
n8n’s pricing page says an execution is a single run of your entire workflow, no matter how many steps it has or how much data it processes. A workflow that loops through 200 invoices in one run is one execution.
As at November 2026, in euros and billed annually, the page lists Starter at €20 a month for 2,500 executions, Pro at €50 for 10,000, and Business at €667 for 40,000 on a self-hosted instance, with Enterprise priced on request and the community edition free to self-host. Overage is capped on Starter and Pro. On Business, workflows keep running if you go over, with possible overage charges. Our guide to n8n pricing goes through each plan.
A worked example: executions against credits
Take a weekly job for a builder. Every Monday it pulls last week’s supplier invoices from the accounts system, matches each to a job, flags any over budget, and posts a summary to the team. Say 150 invoices a week, and five modules that act on each invoice.
- Make: the five per-invoice modules run 150 times each, so about 750 credits a week before the summary steps, or roughly 3,000 to 3,300 a month.
- n8n: one execution a week, about four or five a month, whatever the invoice count.
Now take the opposite shape: a short workflow that fires on every web enquiry and does two things. At 300 enquiries a month that is about 900 Make credits counting the webhook read, against 300 n8n executions. Both fit comfortably on entry plans. The gap opens with record-heavy batch jobs, not with simple triggers.
What Make does well
- The visual map. Scenarios are drawn as connected modules, which makes branches, loops and error paths easy to see at a glance.
- Free routing and error handling. Routers and error handlers cost no credits.
- Graceful stops. Queued webhooks and catch-up polling mean running out of credits delays work rather than losing it.
- Nothing to host. Make runs the platform, so there is no server, update or backup to manage.
What n8n does well
- Batch jobs cost one run. Record-heavy workflows cost the same execution as a two-step one.
- Code when you need it. The Code node runs JavaScript or Python, so awkward data can be fixed in a few lines instead of a chain of modules.
- Your choice of country. A self-hosted instance can sit in an Australian data centre, beside the rest of your systems.
- Clear self-hosting rights. The licence FAQ allows the free community edition for internal business purposes, and allows someone to build and maintain workflows on their own instance for clients who receive only the outputs. It does not allow hosting n8n as a service for others to build on.
Who maintains it, and where the data sits
For an Australian business these two questions often decide it.
Make: Make’s organisation help page says it supports two data centre locations, the United States and the European Union. You choose when you create the organisation and cannot change it afterwards, so a later move means rebuilding in a new organisation. Make handles the platform. Your job is the scenarios themselves, fixing them when a connected app changes.
n8n Cloud: data is stored in Frankfurt, and n8n handles the platform, much like Make.
Self-hosted n8n: you pick the country and the provider, and you take on updates, backups, security patches and monitoring. The n8n hosting docs point to Docker Compose or a one-line setup as the install route. Git version control and separate test and live environments sit on the Business and Enterprise plans.
If either tool will carry personal information, the Privacy Act’s rules on overseas disclosure may apply. Our guide to privacy rules for AI and data tools in Australia covers the principles to check.
When a built and looked-after system fits better
A self-serve tool works well when someone in the business enjoys building scenarios and has time to fix them. It gets harder when automations run invoices, job costing and customer messages, and a quiet failure costs money before anyone notices. That is when having workflows built, monitored and kept current for you can make sense, which is what our automation service does.
It does not suit everyone. If you run a few simple scenarios at low volume, if you or a staff member like owning them, or if a standard Make or n8n Cloud plan does the job, stay self-serve. If you are weighing Make against Zapier instead, our Make vs Zapier guide covers that pairing.