On this page 8
- Zapier, Make and n8n at a glance
- The billing unit is the real price: tasks vs credits vs executions
- Worked example: one 5-step workflow run 1,000 times a month
- Which is easiest to use without a developer?
- Self-hosted n8n: free to download, not free to run
- Adding AI steps to your automations
- How to estimate your cost before you commit
- Zapier, Make or n8n: which should you choose?
The short answer
Zapier is the easiest to start with. Make gives you a flexible visual builder billed in credits. n8n is usually the cheapest for long workflows because it bills once per full run, however many steps it has. The billing unit matters more than the headline price: the same 5-step workflow run 1,000 times a month uses 4,000 Zapier tasks, about 5,000 Make credits or 1,000 n8n executions.
- Translate your workflows into each tool's unit before you compare plan prices.
- Workflows that loop over lists multiply Zapier and Make usage. n8n still counts one execution per run.
- Self-hosted n8n is free to download, not free to run. Someone has to manage the server.
Zapier, Make and n8n at a glance
Prices and limits below were checked on each vendor's pricing page on 2 November 2026. They change often, so confirm them on Zapier's pricing page, Make's pricing page and n8n's pricing page before you buy.
| Tool | What you pay for | Free plan | Entry paid plan | Best for |
|---|---|---|---|---|
| Zapier | Tasks (each successful action) | 100 tasks/month, two-step Zaps only | Pro: $19.99/month billed annually ($29.99 monthly) for 750 tasks | Simple automations, easiest setup |
| Make | Credits (most actions use 1) | 1,000 credits/month | See make.com/en/pricing | Multi-step and branching workflows on a visual canvas |
| n8n | Executions (one full workflow run) | Self-hosted Community Edition, free to download (you run the server) | Cloud Starter: €20/month billed annually for 2,500 executions | Long or high-volume workflows |
If you are new to automation and not yet sure which jobs to automate, start with our guide on how to automate repetitive admin tasks with AI, without writing code. This page assumes you know what you want to automate and need to pick the platform.
The billing unit is the real price: tasks vs credits vs executions
All three tools charge for usage, but each counts usage differently. Zapier counts actions. Make counts credits, which mostly track actions. n8n counts whole workflow runs.
That means the same automation can be cheap on one tool and expensive on another. A $20 plan and a €20 plan are not comparable until you work out how many units your workflows actually use on each. The three definitions below are the key to everything else on this page.
Zapier counts tasks: each successful action
In Zapier, an automation is called a Zap. Every Zap starts with a trigger ("new form entry") and then runs one or more actions ("add a spreadsheet row").
- Counts as a task: each action that completes successfully.
- Does not count: the trigger, filters, Formatter steps and Paths.
So a Zap with a trigger, a filter and two actions uses 2 tasks each time both actions succeed. Usage grows in two directions: more action steps per Zap, and more runs per month.
The Free plan gives you 100 tasks a month, with real limits: two-step Zaps only (one trigger, one action), a 15-minute polling interval for checking new data, and no premium apps. Pro starts at $19.99/month billed annually ($29.99 billed monthly) for 750 tasks. Team is $69/month billed annually ($103.50 monthly) for 2,000 tasks and up to 25 users. Higher task volumes are available at higher prices.
Make counts credits: most actions use 1, some use far more
In Make (formerly Integromat), an automation is called a scenario, built from connected modules. Most module actions use 1 credit. Some AI features use more.
Because usage depends on which modules a scenario contains and how many items it processes, one run can use anywhere from 2 credits to thousands. A scenario that handles one record at a time is predictable. One that processes a large batch can burn through credits quickly.
The Free plan includes 1,000 credits a month. Paid plan prices were not shown on Make's pricing page when we checked, so see make.com/en/pricing for current tiers.
n8n counts executions: one full run, however many steps
In n8n, an execution is one complete run of a workflow, regardless of how many steps (nodes) it contains. A 3-step workflow and a 30-step workflow each use 1 execution per run.
Adding steps does not raise your usage. Running the workflow more often does. That is why n8n tends to be the cheapest option for long, detailed workflows.
On n8n cloud, Starter is €20/month billed annually with 2,500 executions, and Pro is €50/month with 10,000 executions. The self-hosted Community Edition is free to download, with the catches covered further down.
Worked example: one 5-step workflow run 1,000 times a month
Imagine a small business with a website enquiry form. Each new form entry (the trigger) sets off four actions:
- Add a row to a spreadsheet.
- Create a contact in the CRM.
- Send a Slack message to the team.
- Send a welcome email to the person who enquired.
That is 1 trigger plus 4 actions, so 5 steps. It runs 1,000 times a month.
| Tool | How it counts | Usage per run | Usage per month | Fits which listed plan |
|---|---|---|---|---|
| Zapier | Successful actions; trigger is free | 4 tasks | 4,000 tasks | Above Pro (750) and Team (2,000); needs a higher task tier |
| Make | Credits per module | About 5 credits | About 5,000 credits | Above the 1,000-credit Free plan; needs a paid tier |
| n8n | One execution per full run | 1 execution | 1,000 executions | Fits cloud Starter (2,500) |
How the numbers work out
Zapier: the trigger does not count, so each run uses 4 tasks, assuming all four steps are regular actions (not filters or Formatter steps) and all of them succeed. 4 tasks × 1,000 runs = 4,000 tasks a month. That is well above the 750 tasks on Pro and the 2,000 on Team, so you would need to pick a higher task tier on Zapier's pricing page.
Make: if each of the five modules uses 1 credit, that is about 5 credits per run and roughly 5,000 credits a month, around five times the Free plan's allowance. Whether the trigger module itself uses a credit, and whether checks that find no new data use credits, can change the total. Treat this as an estimate and confirm on Make's pricing and help pages.
n8n: each run is 1 execution, so 1,000 executions a month. That fits inside cloud Starter's 2,500 with room to grow.
What changes when a run handles a list of items
The gap widens when a workflow processes several items in one go. Take a second illustration: a daily run that picks up 20 new spreadsheet rows and does 2 actions for each row (say, create an invoice draft and send a notification).
- Zapier: 2 actions × 20 items = about 40 tasks per run. Over a 30-day month, that is about 1,200 tasks.
- Make: each action on each item uses credits, so about 40 credits per run, plus whatever the trigger or search module uses. Again, roughly 1,200 or more a month.
- n8n: still 1 execution per run, so 30 executions a month.
The lesson: loops and bulk processing multiply task- and credit-based costs. Execution-based billing does not grow the same way, because it counts the run, not the items inside it.
Which is easiest to use without a developer?
Price is only half the decision. If nobody on your team can build or fix a workflow, the cheapest tool is not the cheapest choice.
Zapier uses a linear, step-by-step builder: pick a trigger, then add actions one after another. For simple "when this happens, do that" automations, it generally has the gentlest learning curve of the three.
Make uses a visual canvas where modules connect like a flowchart. That makes branching and multi-step logic easier to see and build. It takes a little longer to learn, especially once you start working with lists of items and mapping data between modules.
n8n is node-based and very capable. n8n cloud removes all server work, but the tool still assumes more comfort with data structures, such as the way records and fields are passed between steps. It suits someone who likes spreadsheets and logic more than someone who wants a guided setup.
A practical rule: if the person who will own your automations has never built one, Zapier is the safest first step. Make and n8n reward a bit more learning with more flexibility.
Self-hosted n8n: free to download, not free to run
A common misconception is that n8n is free. The Community Edition is free to download, but you run it on your own server. That means you handle hosting, updates, security patches, backups and fixing it when something breaks.
For a team without technical help, that usually means paying someone to set it up and maintain it, or choosing n8n cloud instead.
Self-hosting n8n: the upside
- No software subscription for the Community Edition itself.
- Full control over where your data is stored and processed.
- Usage is limited by your server's capacity rather than a plan's execution allowance.
- Useful if you have strict rules about keeping data in-house.
Self-hosting n8n: the catches
- You pay for the server and any help to run it.
- Updates, security and backups are your responsibility.
- If the server goes down, your automations stop until someone fixes it.
- Realistically needs someone comfortable managing a server.
Whichever tool you choose, any workflow that moves customer data between apps deserves a privacy check first. Our guide on whether it is safe to put customer data into ChatGPT covers the questions to ask, and they apply to automation platforms too.
Adding AI steps to your automations
All three platforms can call AI models as a step in a workflow. Typical examples: draft a reply to a new customer review for a person to approve (see our human-approved workflow for responding to reviews with AI), or summarise meeting notes and send action items to your task list (see how to choose an AI meeting notes tool).
Budget for AI steps separately:
- Make: some AI features use more than 1 credit per action, so an AI-heavy scenario can use credits faster than the module count suggests.
- All three: calling a model from a provider such as OpenAI or Anthropic usually requires your own account and API key with that provider, billed separately from the automation platform. Some platforms also offer built-in AI options; check how each one is charged on the vendor's page.
Before you automate anything with AI, agree on staff rules for what data may be sent and what needs human review. Our AI acceptable use policy template for small businesses gives you a starting point. For choosing which model to call for writing tasks, see ChatGPT vs Claude vs Gemini for small business writing.
How to estimate your cost before you commit
You can apply the worked example to your own workflows with a pen and paper. The aim is a realistic monthly number for each tool, not a perfect one.
- List the 2-3 workflows you actually need. Ignore nice-to-haves for now; start with the jobs that cost you the most time.
- Count the trigger and action steps in each. Write them down in order, noting any filters or formatting steps.
- Estimate runs per month. If a workflow loops over a list, also estimate how many items each run handles.
- Convert to each tool's unit. Zapier: successful actions × items × runs. Make: roughly credits per module × items × runs. n8n: runs.
- Build your busiest workflow on each free plan or trial. Zapier's Free plan only allows two-step Zaps, so a longer workflow may need a trial or paid plan. After a week, check the usage dashboard and compare it with your estimate.
- Choose the plan with headroom. Then check what happens at the limit: whether runs stop, slow down or trigger extra charges.
Keep your written list of steps. If you ever switch platforms, it makes rebuilding much faster.
Zapier, Make or n8n: which should you choose?
The right tool depends on three things: how long your workflows are, how often they run, and whether you have technical help. Short, occasional automations favour Zapier's simplicity. Long or high-volume workflows favour n8n's per-run billing, and Make sits in between with a flexible visual builder, as long as you keep an eye on credits.
Quick decision
- Short, simple automations, low volume, want the easiest setupZapier.
- Multi-step or branching workflows, comfortable with a visual canvas, can monitor creditsMake.
- Long or high-volume workflows where per-step billing would get expensiven8n cloud.
- Technical help on hand and want full control of data and hostingself-hosted n8n.
- Unsurestart on the free plans and measure your busiest workflow for a week.
Frequently asked questions
Does Zapier count the trigger as a task?
No. According to Zapier's pricing page, only successful actions count as tasks. Triggers, filters, Formatter steps and Paths do not count.
Is n8n really free?
The self-hosted Community Edition is free to download, but you pay for and maintain the server yourself, including updates, security and backups. n8n cloud is paid, starting at €20/month billed annually for 2,500 executions, as checked on 2 November 2026.
What happens when I run out of Make credits?
Your scenarios stop running until credits are available again. Monitor your usage and keep headroom for any automation your customers depend on.
Which tool is cheapest for long workflows with many steps?
Usually n8n, because one execution covers a whole workflow run regardless of how many steps it has. Run your own numbers first, though, because your monthly volume and each plan's limits still matter.
Which is easiest for a complete beginner?
Zapier's linear, step-by-step builder is generally the simplest place to start for basic automations. Make and n8n reward a bit more learning with more flexibility.
Can I move my automations from one tool to another later?
Generally you should expect to rebuild them, because there is usually no direct way to move a workflow between these platforms. Keeping a written list of each workflow's trigger and steps makes switching much faster.
Do I need a developer to use n8n?
Not for n8n cloud, though it suits people who are comfortable with data and logic. Self-hosting n8n realistically needs someone who can manage a server.