n8n charges per workflow execution and can be self-hosted, Make charges per credit (roughly one per module run) and runs only in its cloud, and Zapier charges per task (each successful action) and has by far the largest app catalog. If you want the most apps with the least setup, pick Zapier; if you want visual branching at a low entry price, pick Make; if you have technical staff, long workflows or data that must stay on your own servers, pick n8n.
How do n8n, Make and Zapier compare at a glance?
Every figure below comes from the vendor's own pricing or documentation page, read on 17 October 2026. Prices change often, so treat them as of October 2026 and check the page before you buy.
| n8n | Make | Zapier | |
|---|---|---|---|
| Billing unit | Workflow execution: one full run, any number of steps | Credit: 1 per operation for non-AI apps | Task: 1 per successful action |
| Free option | Free self-hosted Community Edition; cloud trial | Free plan, 1,000 credits a month | Free plan, 100 tasks a month, two-step Zaps |
| Entry paid plan | Starter €20/mo billed annually, 2.5K executions | From $9 for 10,000 credits a month | Professional from $19.99/mo billed yearly |
| Team tier | Pro €50/mo billed annually, 10K executions | Teams plan | Team from $69/mo billed yearly, 25 users |
| Self-hosting | Yes | No, cloud only | No, cloud only |
| Apps and integrations | 2,300+ listed, plus HTTP requests and code | 3,000+ native apps | 9,000+ apps |
| AI agents | AI Agent node on every plan | Make AI Agent (open beta) | Zapier Agents, separate activity quota |
| Data location | Cloud: Frankfurt; self-hosted: wherever you run it | US1, US2, EU1 or EU2 data zone | AWS US-East 1 |
Sources: n8n pricing, n8n vs Zapier, n8n integrations, Make pricing, Make vs n8n, Make data zones, Zapier pricing, Zapier data residency.
How does each tool count usage?
This is the question that decides your bill, and each vendor answers it differently.
n8n counts executions. Its pricing page says "An execution is a single run of your entire workflow. It doesn't matter how many steps are in the workflow or how much data it processes." A 3 step workflow and a 40 step workflow cost the same per run, and a run that processes 500 rows is still one execution.
Make counts credits. Credits replaced operations as Make's billing unit, and for ordinary apps "1 operation equals 1 credit." An operation is "a single module run to process data or check for new data", and modules run once per bundle, so a trigger that returns 10 records makes every later module run 10 times. Make's own example: a trigger returning 10 form responses followed by three modules uses 31 operations in one scenario run. AI modules can use a varying number of credits based on tokens.
Zapier counts tasks. "Each successful action in a Zap counts as a separate task." Triggers are free, and so are Filter, Paths, Formatter, Delay, Looping and several other built-in steps, according to Zapier's task usage guide. Some products use more: a Zapier MCP tool call uses 2 tasks, and AI by Zapier steps depend on the model tier you pick.
Here is the same workflow priced in each unit: a trigger plus five actions, handling one record per run, 1,000 runs a month.
| n8n | Make | Zapier | |
|---|---|---|---|
| Per run | 1 execution | 1 trigger operation + 5 = 6 credits | 5 tasks (trigger free) |
| Per month | 1,000 executions | 6,000 credits, plus 1 credit per empty polling check | 5,000 tasks |
The gap widens with longer workflows and with batches. Add ten more steps and n8n still counts 1,000 executions, while the Make and Zapier numbers roughly triple.
How much do n8n, Make and Zapier cost?
As of October 2026, from each vendor's own pages:
n8n. Starter is €20 a month billed annually for 2.5K executions with 5 concurrent executions. Pro is €50 a month billed annually for 10K executions. Business is €667 a month billed annually for 40K executions and is currently self-hosted only. Enterprise is priced by sales. All plans "include unlimited users & workflows and every integration." Business overages are billed at 4,000 EUR per extra bucket of 300,000 executions. The page shows prices in euros.
Make. The pricing page lists Free, Core, Pro, Teams and Enterprise plans and states that the Free plan includes 1,000 credits per month. Make's own comparison page says paid plans start "at $9 for 10,000 credits per month", and the pricing page notes that annual payment is cheaper than monthly. Extra credits can be bought in blocks when you run out.
Zapier. The Free plan is $0 with 100 tasks a month and two-step Zaps. Professional starts from $19.99 a month and Team from $69 a month when paid yearly, which the page labels as a 33% saving. Each plan has task tiers, and Zapier says "The higher the task tier, the lower the cost per task." If you hit your limit with pay-per-task billing on, runs continue at a higher per task rate; with it off, Zaps pause until the next period.
Is Make or n8n cheaper? For short workflows at low volume, Make's entry price is lower. For long workflows, loops over many records, or anything running thousands of times a month, n8n's per execution unit usually wins, and self-hosting removes the subscription entirely at the cost of running a server.
Can you self-host n8n, Make or Zapier?
Only n8n. n8n's comparison page describes a "free Community Edition with a fully featured workflow editor", and its documentation explains how to self-host n8n on your own infrastructure, on premises or in a private cloud, using Docker Compose or other methods. Read the license before you build a product on it: the Sustainable Use License allows use "only for your own internal business purposes or for non-commercial or personal use", and anything else needs a commercial agreement with n8n.
Zapier describes itself as "a cloud-based AI orchestration platform", and Make's own comparison table calls Make "Cloud-first at all subscription levels". Self-hosting is not free in practice either: you take on setup, updates, uptime and security, which is the trade off Zapier's comparison page points to.
Which has better AI agent features?
All three now ship agents, built differently.
n8n's documentation says you can "connect different LLM providers such as OpenAI, Anthropic, and Google, add tools and memory, and combine several models in one workflow." The pricing page lists agents, MCP and AI nodes on every cloud plan. Because agent runs sit inside an ordinary workflow, each run is still one execution.
Make released the current version of its agent app on February 2, 2026, and its documentation says the Make AI Agent (New) app is in open beta, so "product functionality and pricing may change." Make's own guidance is sensible: choose tasks for an agent that you would trust an intern to handle, and avoid high stakes financial decisions.
Zapier sells Agents as a separate product. Agents usage does not count as tasks and draws on its own activity quota: the pricing page lists 400 activities a month on Agents Free and 1,500 on Agents Pro at $33.33 a month billed annually.
Where is your data stored?
This matters if you handle customer data under GDPR or client contracts.
- Zapier says "Customer data is stored in AWS US-East 1 (United States)" and that region-specific residency is not available unless otherwise agreed.
- n8n Cloud stores data "within the EU, on servers located in Frankfurt, Germany"; self-hosted n8n stores it wherever you host it.
- Make offers US1, US2, EU1 and EU2 data zones, but its documentation warns that an EU zone "does not automatically mean every request gets processed and stays in the EU": some AI features, including all OpenAI models in Make's AI Provider, always route to the US.
Which should you choose?
| Your situation | Best fit | Why |
|---|---|---|
| Non-technical team, many SaaS apps, simple flows | Zapier | Largest catalog and the least setup |
| Visual branching on a small budget | Make | Low entry price, data zone choice |
| Long workflows or high volume | n8n | One execution per run regardless of steps |
| Data must stay on your servers | n8n self-hosted | The only one of the three you can run yourself |
| You want to embed automation in a product you sell | Check licenses first | n8n's free license excludes it; the cloud tools need their own terms |
In practice the tool is rarely the hard part. The hard part is writing down the trigger, inputs, decisions and owner of the process, which we cover in our guide to business process automation. Any of these three can carry a well mapped process, and when a team wants it designed, built and maintained for them, that is the kind of workflow automation we scope. For CRM heavy work, see CRM automation, and if you run a small team, our guide to AI automation for small business covers where to start.
Frequently asked questions
Is n8n free?
The self-hosted Community Edition costs nothing to license, within the limits of the Sustainable Use License.
You still pay for the server you run it on and the time to maintain it. n8n Cloud and the Business and Enterprise plans are paid.
Can Zapier replace n8n?
For most connections between popular SaaS apps, yes, and with less setup.
It cannot replace n8n where you need self-hosting, custom code nodes that run on your own infrastructure, or very long workflows billed per run rather than per step.
Is Make easier than Zapier?
Zapier is usually easier to start with, because a Zap is a straight list of steps.
Make's canvas shows branches and loops visually, which helps once a workflow has several paths, but its bundle and operation logic takes longer to learn.
Do triggers use tasks or credits?
On Zapier, no: triggers and polling never use tasks.
On Make, yes: a trigger module uses one operation each time it checks for new data.
On n8n, the trigger simply starts the run, and the whole run counts as one execution.