If your automations are simple and low-volume, Zapier's per-task billing is fine and the cheapest thing about it is your time. The moment a single workflow fans out into many steps, or you start running thousands of runs a month, the pricing model — not the sticker price — is what decides your bill. Make charges per module execution, which is finer-grained than a Zapier task; n8n charges per workflow execution regardless of how many steps that workflow has, and it can be self-hosted for infrastructure cost only. The switch point is almost always about billing units, not features.
I've run all three in production for internal automations, and the migrations I've done were never triggered by a missing feature. They were triggered by a monthly invoice that grew faster than the value of the work being automated. This post is about spotting that inflection before the invoice does.
How does each tool actually count usage?
The three tools use three different meters, and conflating them is where most cost surprises come from.
Zapier bills per task. A task is one action step that successfully runs. The trigger that starts a Zap does not count; every action after it does. So a Zap that watches a form and does one thing costs one task per submission. A Zap that watches a form, looks up a record, formats a value, and writes to two places costs four tasks per submission. Filters and paths that stop early generally don't consume a task, which matters more than people expect.
Make bills per operation. An operation is a single module doing a single unit of work. It's conceptually similar to a Zapier task, but Make's modules are more granular and the included volumes on comparable tiers are typically much higher, so the effective cost per unit of work tends to be lower. The catch is that iterators, aggregators, and array-processing modules can multiply operations fast — a scenario that loops over 50 items can spend 50+ operations in one run.
n8n bills per execution. One workflow run is one execution, whether that workflow has three nodes or thirty. This is the structural difference that changes everything at scale: a workflow's complexity is decoupled from its cost. n8n is also fair-code / source-available (Sustainable Use License, not OSI-open), and the self-hosted edition has no execution meter at all — you pay for the server it runs on.
The takeaway: Zapier and Make charge roughly by how much work happens inside a run, while n8n charges by how many runs happen.
What does that mean for a multi-step workflow?
Take one realistic workflow: a new lead comes in, you enrich it, score it, write it to your CRM, post to a Slack channel, and log it to a sheet. That's five actions after the trigger.
| Meter | Cost per single run of this workflow | Scales with |
|---|---|---|
| Zapier (per task) | 5 tasks | Number of steps × number of runs |
| Make (per operation) | ~5–6 operations (more if any step loops) | Number of modules × number of runs, loops multiply |
| n8n (per execution) | 1 execution | Number of runs only |
Now multiply by 2,000 leads a month. Zapier meters 10,000 tasks. Make meters a similar order of magnitude in operations, but usually with a much larger included allowance per dollar. n8n meters 2,000 executions — the same 2,000 you'd pay whether the workflow had five steps or fifteen.
This is why "just count your triggers" is bad advice. On per-task billing, a workflow with ten steps is ten times more expensive than a one-step workflow running the same number of times. Your bill is a function of workflow depth, and depth is exactly what grows as automations mature.
The takeaway: on per-task or per-operation billing, adding steps to an existing workflow silently raises your recurring cost; on per-execution billing, it doesn't.
When is Zapier still the right call?
Don't let the pricing analysis talk you out of the tool that saves you the most time. Zapier's real product is its app catalog and reliability — it has the broadest set of maintained integrations of the three, and its Zaps tend to "just work" against flaky third-party APIs because Zapier absorbs a lot of edge-case handling for you.
Zapier stays worth it when:
- Your total monthly task volume is modest and predictable.
- You value not maintaining anything — no server, no updates, no debugging someone else's self-hosted instance.
- You depend on a niche integration that Make or n8n don't cover as cleanly.
- The people building automations are non-engineers who'll be most productive in Zapier's guided editor.
Where it stops making sense is high-volume, high-step-count automation, especially anything that loops over collections. That's the scenario where task math turns against you fastest.
The takeaway: Zapier is priced for convenience, and convenience is genuinely worth paying for until volume makes the per-task meter the dominant line on your bill.
When does Make win the middle ground?
Make is the tool I reach for when a workflow is genuinely complex — branching, iterating over arrays, aggregating results — but I still don't want to run infrastructure. Its visual scenario builder handles data transformation and multi-path logic more expressively than Zapier's linear editor, and its per-operation allowances usually give you more headroom per dollar at moderate scale.
The honest limitation: operations are easy to under-estimate. A scenario that fetches a list and processes each item can burn through operations far faster than a napkin estimate suggests, because every iteration is metered. You have to design with the meter in mind — batching, filtering early, and avoiding needless module calls. Make also has a steeper learning curve than Zapier; the power comes with more concepts to hold in your head.
The takeaway: Make is the value pick for complex-but-managed automation, as long as you actually model your operation count instead of guessing it.
When is n8n worth running yourself?
n8n becomes compelling at exactly the point the other two get expensive: high volume, deep workflows, or both. Per-execution billing means a twenty-node workflow costs the same per run as a two-node one, and self-hosting removes the execution meter entirely in exchange for running a container.
That trade is real, though, and I won't pretend it's free. Self-hosting n8n means you own uptime, upgrades, backups, and security patching. When a workflow breaks at 2 a.m., there's no vendor support line — it's you and the logs. n8n Cloud exists precisely so you can get per-execution pricing without the ops burden, and for many teams that's the sweet spot. n8n's integration catalog is narrower than Zapier's, but it has strong HTTP and code nodes, so anything with an API is reachable even without a prebuilt connector.
Use n8n when you have the engineering capacity to either run it or reason about it, and when your volume is high enough that decoupling cost from workflow complexity actually saves meaningful money.
The takeaway: n8n's per-execution model is the one that doesn't punish you for building sophisticated workflows — but self-hosting swaps a software bill for an operational responsibility.
A quick decision guide
| Your situation | Best fit |
|---|---|
| Low volume, non-technical builders, want zero maintenance | Zapier |
| Complex branching/looping logic, still want a managed tool | Make |
| High volume and/or deep multi-step workflows, some engineering capacity | n8n (Cloud or self-hosted) |
| Cost must decouple from workflow complexity | n8n |
| Need the widest catalog of maintained integrations | Zapier |
Bottom line
Pick Zapier when the value of your time and its integration breadth outweigh the per-task meter — that's most small teams starting out. Move to Make when your workflows get genuinely complex but you still don't want to run infrastructure, and budget your operations honestly because loops multiply them. Reach for n8n when volume or workflow depth makes per-step billing the biggest line on your invoice, and you have the capacity to self-host or pay for n8n Cloud's per-execution model. The signal to switch is never a missing feature — it's the day your bill grows faster than the work it automates.
Top comments (0)