Most people don't leave Zapier because it stops working. They leave because the task meter does. Make, n8n and Pabbly Connect all run similar automations, but each one counts usage differently — and that counting rule, not the headline price, decides what the same workflow costs you over a year.
Core, billed annually, with 10,000 credits/month. Free plan: 1,000 credits/month.
Cloud Starter, billed annually, 2,500 executions/month. Or self-host the Community edition for no licence fee.
Standard, billed annually, 10,000 tasks/month. One-time lifetime plans are also published.
Professional, billed annually, from 750 tasks/month. Free plan: 100 tasks/month, two-step Zaps only.
A "task", a "credit" and an "execution" are not three names for the same unit. They are three different meters, and the meter is the product you're really buying:
Every successful action step in a Zap counts as one task. The trigger is free. A five-action Zap that runs 1,000 times uses roughly 5,000 tasks. Some AI steps consume more than one task depending on the model used.
Every module that runs in a scenario costs roughly one credit (the unit was renamed from "operations" in 2025, converted 1:1). A five-module scenario running 1,000 times uses roughly 5,000 credits — but the entry paid plan includes 10,000 of them.
One run of a whole workflow counts as one execution, whether it has two steps or forty. Step count is simply not part of the bill on Cloud plans — and self-hosting has no vendor meter at all.
Actions that call an outside app count as tasks. Internal steps — filters, routers, formatters — do not. Two workflows that look identical on a diagram can therefore meter differently here than on Zapier.
The practical consequence: before comparing prices, count two numbers for your own automations — how many times they run per month, and how many action steps an average run has. Every price difference below follows from those two numbers.
Take one fixed workload and price it everywhere: 2,000 workflow runs a month, averaging 5 action steps per run. That is 10,000 billable units on the per-step meters, and still just 2,000 executions on n8n. These are calculations from published prices checked in November 2026 — arithmetic, not a benchmark, and not a promise about your bill.
| Platform | Units this workload consumes | Published plan that covers it | Cost of that plan |
|---|---|---|---|
| Zapier | 10,000 tasks | Professional, 10,000-task tier | $129/mo billed annually = $1,548/year |
| Make | ~10,000 credits | Core (10,000 credits included) | $9/mo billed annually = $108/year |
| n8n Cloud | 2,000 executions | Starter (2,500 executions included) | €20/mo billed annually = €240/year |
| n8n self-hosted | 2,000 executions | Community edition (no licence fee) + a small server, e.g. a published 2 GB VPS plan at $12/mo | $144/year in server cost — plus your own maintenance time, which no price table can quote for you |
| Pabbly Connect | Up to 10,000 tasks (fewer if some steps are internal) | Standard (10,000 tasks included) | $16/mo billed annually = $192/year |
When the ranking flips. Stretch the same 2,000 runs to an average of 10 steps and Zapier's tier rises to $189/mo billed annually ($2,268/year), while n8n Cloud stays at €20/mo — its meter never saw the extra steps. Go the other way — many runs, one or two steps each — and run count becomes the binding constraint instead: n8n Cloud's 2,500-execution Starter ceiling arrives quickly, while Make's 10,000-credit allowance absorbs short workflows comfortably. Neither platform is "cheaper" in the abstract. Each is cheaper for a different shape of work.
Assumptions, stated so you can redo the maths with your own numbers: Make module counts are approximated as equal to action-step counts; Pabbly tasks are counted conservatively as if every step were an external action. EUR and USD figures are shown side by side without conversion, because n8n publishes in EUR and the rate you'd actually pay depends on your billing region.
| Make | n8n | Pabbly Connect | Zapier | |
|---|---|---|---|---|
| Best for | Visual builders whose Zapier task bill outgrew the work | Technical users, step-heavy workflows, self-hosters | Budget-fixed solo users and small businesses | Teams that need a specific app integration, or are just starting |
| What it counts | Credits — roughly 1 per module run | Executions — 1 per workflow run, any length | Tasks — external actions only; internal steps are free | Tasks — 1 per successful action step |
| Free option | 1,000 credits/mo, 2 active scenarios | Community edition: free to self-host, no vendor execution limit | 100 tasks/mo | 100 tasks/mo, two-step Zaps only |
| Entry paid plan | Core $9/mo billed annually — 10,000 credits | Cloud Starter €20/mo billed annually — 2,500 executions | Standard $16/mo billed annually — 10,000 tasks | Professional $19.99/mo billed annually — 750 tasks |
| How the price grows | Same 10,000-credit start on Core / Pro ($16) / Teams ($29); larger credit bundles cost more | Pro €50/mo for 10,000 executions; Business (self-hosted licence) jumps to €667/mo for 40,000 | Unlimited-tasks tier at $79/mo billed monthly ($69/mo billed annually); or one-time lifetime plans | Steep task ladder: 2,000 tasks $49/mo, 5,000 $89/mo, 10,000 $129/mo (all billed annually) |
| Published app count | 3,000+ | 500+ native integrations, plus an HTTP node for any API | 2,000+ | 8,000+ |
| Self-hosting | No | Yes — a defining feature, under its Sustainable Use Licence | No | No |
| Trade-off to know | The visual builder is more powerful and less immediately obvious than Zapier's; complex scenarios take planning | Self-hosting means you own updates, backups and uptime; the Business licence tier is a large price jump from Pro | Smaller app catalogue and a less polished builder than Zapier; lifetime plans carry vendor-longevity risk and varying support terms | The task ladder is the most expensive way here to run step-heavy or high-volume work |
Prices are before tax, as published by the vendors in November 2026. n8n prices are shown in EUR as the vendor publishes them; your displayed currency and price can differ by region. App counts are the vendors' own published figures and count different things (apps, integrations, nodes), so treat them as a rough sense of catalogue size, not a like-for-like score. Plans change — always confirm the current price and what's included on the vendor's site before you buy.
Make (formerly Integromat) builds automations as visual scenarios: modules on a canvas, with routers, iterators, aggregators and error handlers as first-class pieces. Its pricing is where most Zapier switchers land first. The free plan includes 1,000 credits a month (two active scenarios, 15-minute minimum interval), and every paid tier starts from the same 10,000 credits: Core at $9/mo billed annually, Pro at $16/mo, Teams at $29/mo, with larger credit bundles priced above that. Because a credit is roughly one module run, the honest comparison against Zapier is allowance size: 10,000 credits at $9 against 750 tasks at $19.99. The catch is symmetrical — a sprawling scenario with iterators can burn credits faster than its diagram suggests, so the meter still rewards tidy design.
n8n prices a completely different unit: one execution is one run of an entire workflow, regardless of how many nodes it contains. On n8n Cloud, Starter is €20/mo billed annually for 2,500 executions and Pro is €50/mo for 10,000, with unlimited users and active workflows. The Community edition can be self-hosted for no licence fee under n8n's Sustainable Use Licence — your real costs become the server (small published VPS plans start around $12/mo) and your own time for updates, backups and keeping it online. Two cautions belong in any honest summary: the self-hosted Business licence jumps to €667/mo billed annually, and self-hosting is an operations commitment, not a free lunch. Its native catalogue (500+ integrations) is the smallest here, partly offset by a capable HTTP node and code nodes that can reach almost any API — if you're willing to build the connection yourself.
Pabbly Connect competes almost entirely on its meter and its price list. It counts a task only when an action calls an outside app — filters, routers and formatters run free — so workflows padded with internal logic meter lower here than on Zapier. Subscriptions start with a free 100-task plan, then Standard at $16/mo billed annually for 10,000 tasks, up to an unlimited-tasks tier at $79/mo billed monthly. Unusually, it also publishes one-time lifetime plans: $349 for 3,000 tasks/month, $799 for 10,000, and $1,298 for 20,000. Read those carefully before treating them as a bargain: the cheapest lifetime tier carries workflow and step limits, support terms differ by lifetime plan, and any lifetime deal is a bet that the vendor keeps the product running on today's terms for years. On simple maths, the $799 tier equals roughly 4.2 years of the Standard subscription — whether that is a good trade depends on the risk, not just the division.
Zapier remains the reference point: the broadest published app catalogue (8,000+), the most templates and outside tutorials, and the builder most non-technical teammates can pick up unaided. Its pricing is the reason this page exists. The free plan allows 100 tasks a month and two-step Zaps only. Professional starts at $19.99/mo billed annually for 750 tasks, and the ladder climbs by task volume: 2,000 tasks at $49/mo, 5,000 at $89/mo, 10,000 at $129/mo, 20,000 at $189/mo, all billed annually, with Team plans from $69/mo. None of that makes Zapier a bad product — at a few hundred tasks a month it is often the rational choice, and for an app only Zapier integrates, it is the only choice. The problem is purely structural: because every action step meters, success is what gets expensive. The more your automations do per run, and the more they run, the faster you climb a ladder your alternatives price very differently.
From your current tool's usage page: runs per month, and average action steps per run. Multiply them. That product is your Zapier / Make / Pabbly volume; the run count alone is your n8n volume. Every decision below is just pricing those two numbers.
List the apps your automations actually touch and confirm each one exists on the platform you're considering. A cheaper meter is worth nothing for an integration that isn't there — this is the one check that most often sends people back to Zapier, legitimately.
Few runs, many steps → n8n's per-execution model. Many runs, few steps → Make's large credit allowance. Logic-heavy workflows full of filters → Pabbly's free internal steps. Small volume, maximum coverage → staying on Zapier is a defensible answer, not a failure.
Make's free plan, Pabbly's free plan, a self-hosted n8n install and Zapier's own usage page all let you measure instead of guessing. Rebuild your busiest automation, run it for a few days, and read the meter. Then decide with your numbers, not this page's examples.
Because the starting price buys the starting task tier — 750 tasks a month on Professional. Every action step in every Zap run draws from that pool, so a handful of five-step Zaps running hourly can pass 750 tasks quickly, and the next tiers cost $49, $89 and $129 a month billed annually. The plan didn't change; the meter ran. Your usage page shows exactly which Zaps consumed what.
A Zapier task is one successful action step. A Make credit is roughly one module run — a similar per-step idea, but sold in much larger entry allowances. A Pabbly task is one external action, with internal steps excluded. An n8n execution is one full run of a workflow at any length, so it is the only unit here that ignores step count entirely. Same automation, four different bills — that is the whole comparison in one paragraph.
The Community edition carries no licence fee and no vendor execution limit. It is not cost-free in total: you pay for a server (small published VPS plans run about $12–24/mo), and you take on updates, backups, security and uptime yourself. Teams that later need the Business features (SSO, multiple environments, Git version control) face a self-hosted licence from €667/mo billed annually. Price the whole arrangement, not just the licence line.
Arithmetically, the $799 lifetime tier (10,000 tasks/month) equals about 4.2 years of the Standard subscription at $16/mo billed annually — less time if the alternative in your case is a higher Zapier tier. But a lifetime plan prepays a single vendor for years of service: allowances and support terms differ from the subscription tiers, and the product's future terms are the vendor's to change. It suits buyers who have read the specific tier's limits and accept that risk knowingly.
More often than alternative-comparison pages admit. If your task volume is low, if a tool you depend on is only integrated with Zapier, or if your team's time costs more than the price gap, staying put is the economical choice. The case for switching gets strong when step-heavy workflows run at volume — exactly the pattern the workload table above prices out.
No — and we won't pretend otherwise. This guide is compiled from publicly available vendor pricing and specifications, checked in November 2026, plus arithmetic on those published figures with the assumptions stated next to each calculation. Where a judgment appears, it follows from the vendors' own published meters and prices, which you can verify on their pricing pages before spending anything.