Most comparisons of these three are written from the vendors' feature lists. We build automations on all three, so this one starts from the questions that decide the choice: how often will it run, who will maintain it, how sensitive is the data, and does anyone need to write code. Every price and limit below comes from the vendors' own pages.
The quick verdict
Choose Zapier for the fastest route to a working automation and the widest app coverage, if your volumes are modest. Choose Make if your workflows branch, loop and process lists, and you want a low price per step. Choose n8n if you need custom code, high volumes or control over where data is hosted, and you have someone to maintain it.
| Tool | Best for | Watch out for |
|---|---|---|
| Zapier | Non-technical teams, simple app-to-app workflows, the biggest app catalogue | Cost rises quickly as every successful action step counts as a task |
| Make | Visual, multi-branch workflows and data handling at a low price per step | Credits are used by nearly every module, so long scenarios add up |
| n8n | Custom code, heavy volumes, self-hosting and data control | Someone has to build and maintain it properly, and the licence has limits |
Who this is for
This is written for owners and operations leads at small and mid-sized businesses connecting tools like HubSpot, Xero, Slack, Gmail and a CRM, and for whoever will be asked to maintain the result. If you plan to build automation into a software product you sell, the n8n licence section matters most.
The three contenders
Zapier is a hosted service. You build "Zaps" from a trigger and actions. It states that it connects more than 10,000 apps, the largest catalogue of the three.
Make is also a hosted service. You build "scenarios" on a visual canvas. Its pricing page lists 3,500+ apps, and hosting is on AWS in the EU or North America on every plan.
n8n is the odd one out. You can use n8n's hosted cloud, or you can run the Community edition yourself, which n8n's pricing page says is available on GitHub. It is released under the Sustainable Use License, a source-available licence that n8n describes as fair-code rather than open source. Its integrations page advertises 2,000+ integrations.
Pricing: how each one charges
The sticker price matters less than the unit each tool counts. Zapier counts tasks, Make counts credits and n8n counts executions. They are not the same size of unit, and that is where most cost surprises come from.
| Pricing | Zapier | Make | n8n |
|---|---|---|---|
| Free tier | 100 tasks/month, two-step Zaps, 15-minute polling | 1,000 credits/month, 2 active scenarios, 15-minute minimum interval | No hosted free plan. Cloud trial with 1,000 executions, no card needed. Community edition is free to self-host |
| Entry paid plan | Professional from $19.99/month annual ($29.99 monthly) for 750 tasks | Core $9/month at 10,000 credits, annual billing | Starter €20/month billed annually, 2.5K executions |
| Next step up | Team from $69/month annual (2,000 tasks), 25 seats | Pro $16, Teams $29 per month at 10,000 credits, annual billing | Pro €50/month annual, 10K executions. Business €667/month, 40K executions, self-hosted only |
| What counts as usage | A task is a successful action step. Triggers, Filter, Paths and Formatter steps are not counted | A credit is a module action. Routers and error handler modules use none | An execution is one run of the entire workflow, however many steps it has |
| Self-hosting | Not listed, hosted by Zapier | Not listed, hosted on AWS | Yes, free Community edition |
| Currency shown | USD | USD | EUR |
Prices checked October 2026 on zapier.com/pricing, make.com/en/pricing and n8n.io/pricing. None of the three shows GBP, so figures are quoted as shown in USD or EUR, and they change often.
The counting rules are in each vendor's documentation. Zapier's task usage article says "a task is any successful action that runs in Zapier" and that triggers never use tasks. Make's pricing FAQ says most scenario actions use one credit, with routers and error handlers free. n8n's pricing page says "an execution is a single run of your entire workflow."
Take a workflow with five action steps that runs 1,000 times a month. On Zapier that is about 5,000 tasks. On Make it is at least 5,000 credits. On n8n it is 1,000 executions. Same workflow, three very different meters. At the 5,000-task tier Zapier lists Professional at $89 a month on annual billing, about $17.80 per 1,000 tasks, while Make lists Core at $0.90 per 1,000 credits.
Zapier's free plan is limited to two-step Zaps, and multi-step Zaps arrive with Professional.
Features and capabilities compared
| Capability | Zapier | Make | n8n |
|---|---|---|---|
| Stated integrations | More than 10,000 apps | 3,500+ apps | 2,000+ integrations |
| Branching and logic | Paths on paid plans: up to 10 branches per group, 3 nested levels, one fallback branch | Routers and filters on every plan | If node for a two-way split, Switch node for multiple routes |
| Code steps | JavaScript (Node.js 22) and Python 3.13. Extended runtimes of up to 10 minutes on paid plans | Code app for JavaScript or Python on paid plans, at 2 credits per second of run time | Code node for JavaScript and Python. Self-hosted can import npm modules, cloud cannot |
| AI and agent features | AI by Zapier steps count as tasks. Zapier Agents usage is tracked separately | AI agents (beta), AI toolkit and MCP server on all plans | AI Agent node with chat models, tools, memory and MCP servers |
| Where your data sits | Zapier states it stores and processes personal data in the US | Choose a US or EU data centre when you create the organisation. It cannot be changed later | Managed cloud under n8n's data processing agreement, or self-host and choose the location yourself |
| Users | 1 seat on Free and Professional, 25 on Team | Unlimited users on every plan | Limits on shared projects by plan, and the Community edition has no projects feature |
Sources for the table: Zapier's apps page, Paths and Code help articles and data transfer assessment; Make's pricing page and organisations article; n8n's Code node, AI and Switch docs.
One caution: the vendors define an "integration" differently, so the counts are not directly comparable. What matters is whether the apps you use have the specific actions you need, so check that before committing.
Data location and GDPR for UK businesses
If you handle customer personal data, where the workflow engine runs is a real decision. Zapier says it stores and processes personal data in the US. Make offers a US or EU data centre, fixed when you create the organisation, so neither region is the UK. n8n is the only one where you can choose the server location yourself, because on a self-hosted install, as n8n's privacy documentation puts it, n8n is neither a controller nor a processor of your data.
That control comes with work: deleting user data, patching, backups and uptime become yours. For many UK small businesses a managed tool with a signed data processing agreement is the lower-risk option. The bigger risk is usually sending a full customer record through four steps when two fields would do.
The n8n licence, in plain English
n8n's Community edition runs under the Sustainable Use License. It allows use for your own internal business purposes or non-commercial use. n8n's licence FAQ gives the practical reading: running it to help run your business is fine, and so is building automations for clients on your instance as long as they cannot create or edit them. Hosting n8n as a service so others can build their own workflows is not permitted without a commercial agreement. For a business automating its own processes, none of this gets in the way.
What each one is actually best for
Zapier is best for teams who want something working this afternoon. The app catalogue is the biggest and the editor the gentlest. It suits low-volume, linear workflows such as "new form entry, create contact, send Slack message." The cost of that ease shows up when volumes grow or workflows get long.
Make is best for workflows that process data rather than just pass it along: looping through order lines, splitting by condition, combining results, and handling errors on purpose. The canvas shows what the workflow does, which helps when someone else inherits it, at a low credit price.
n8n is best for teams with an unusual API to call, real code to run, a lot of volume, or data that has to stay on infrastructure you control. It rewards technical care, and a self-hosted instance nobody updates becomes a liability.
How we actually decide
We build automations on all three, connecting tools such as HubSpot, Xero, Slack, Gmail and client CRMs, and we have no house default. Each project goes through the same short list of questions, and the tool falls out of the answers.
- How many times will it run, and how many steps per run? Multiply the two. Low volume and short workflows suit Zapier, long branching ones suit Make, and very high volume points to n8n because it charges for the run, not the step.
- Who will maintain it? If it is someone non-technical on your team, pick the tool they can open and fix, usually Zapier or Make. If we maintain it under a retainer, n8n becomes viable because we handle updates and monitoring.
- How sensitive is the data? Health, legal, financial or children's data pushes us toward a deployment we can place and document: n8n self-hosted, or Make in the EU region with minimal personal data passing through.
- Does it need custom code or an unusual API? One small data tweak is fine in any of the three. If the workflow is mostly code, n8n is the more natural home.
- How fast does it need to ship? A quick test of whether a process is worth automating is fastest in Zapier. We would rather prove the value there and rebuild properly than over-engineer something nobody uses.
If your workflows are mainly about keeping a CRM in step with the rest of your stack, our CRM integration service covers exactly that. Our piece on whether your website is ready for AI agents is a useful companion.
Frequently asked questions
Is n8n cheaper than Zapier?
Usually, once your workflows get busy. Zapier bills per successful action step, so a five-step workflow run 1,000 times uses about 5,000 tasks. n8n bills per execution, meaning one run of the whole workflow, so the same workload is 1,000 executions however many steps it has. Self-hosting the Community edition has no licence fee, but you pay for the server and the time to maintain it.
Is Make cheaper than Zapier?
Per unit, yes by a wide margin. At the time of checking, Make's Core plan was listed at $0.90 per 1,000 credits on annual billing, while Zapier's Professional plan works out at about $17.80 per 1,000 tasks at the 5,000-task tier on annual billing. The units differ, though: Make counts steps that Zapier does not, so compare the cost of one real workflow.
Which is easiest for a non-technical team: Zapier, Make or n8n?
Zapier is the quickest to get a first workflow running. Make suits people who think in flows and can cope with a steeper learning curve. n8n gives the most control but expects someone comfortable with data structures, expressions and, if self-hosted, servers.
Is n8n really free?
The self-hosted Community edition is free to run, under n8n's Sustainable Use License, which allows use for your own internal business purposes. It is source-available rather than open source. Features such as SSO, environments and projects need a paid plan, and n8n's hosted cloud plans are priced separately.
Can a UK business use Zapier, Make or n8n and stay on top of GDPR?
Yes, but where your data sits differs. Zapier states it stores and processes personal data in the US. Make lets you choose a US or EU data centre when you create an organisation, and you cannot change it afterwards. n8n Cloud comes with a data processing agreement, and a self-hosted n8n install lets you choose the server location yourself, including the UK. Whichever you pick, minimise the personal data passing through each workflow.
Can we start on one tool and move to another later?
Yes, but plan for a rebuild rather than an export, because a Zap does not convert into a Make scenario or an n8n workflow. Documenting each workflow in plain English makes the move far less painful.
Where to start
Write down the three workflows costing you the most time, estimate how often each runs and how many steps it has, and check those numbers against the tables above. If you would rather have someone who builds on all three make the call, our AI automation service starts with exactly that question, or you can talk to the team and get a straight answer.
Jinnat Ul Hasan
Founder & CEO, Whizz People
