What will Zapier, n8n and Make actually cost you?
Most cost calculators ask how often your automation runs, multiply by a headline price, and hand you a number. That number is wrong, because these tools do not sell the same thing. A Zapier task is one step. An n8n execution is one whole run. If a calculator never asks how many steps you have, it cannot tell those apart.
This one asks. Same three inputs everyone else uses, plus the one they skip.
We make a competing product.BizOps Pilot is in the table below, it is ours, and it charges a flat price with no meter. Read our row with that in mind. We have tried to earn the rest of the page: every competitor number here was read off that vendor's own pricing page or help centre on 2026-07-16 and links back to it, we left out every figure we could not verify at the source rather than guessing it, and where a competitor beats us we say so in their row. n8n self-hosted, for one, is cheaper than we are and gives you the same no-meter deal.
The method, in one table
Three vendors, three different units, one workload. This is the entire trick.
| Tool | Unit | Which means | Do steps multiply it |
|---|---|---|---|
| Zapier | task | one successful action stepA 3-step workflow run 500 times is 1,500 tasks. | yes |
| n8n | execution | one whole workflow run, any number of stepsThe same workflow run 500 times is 500 executions. | no |
| Make | credit | one module action, and more for AIThe same workflow run 500 times is 1,500 credits, at minimum. | yes |
| BizOps Pilot | none | there is no unitThe same workflow run 500 times, or 500,000 times, is one flat price. | no |
Read that Zapier row and that n8n row together. Same automation, same schedule, three times the billable units. That gap is not a price difference. It is a units difference, and it is invisible on a pricing page.
Your workload
Four numbers. The third one, steps per workflow, is the one every other calculator leaves out.
Distinct automations, not runs.
How often the trigger fires.
The trigger does not count. Nor do filters or paths. Zapier charges for none of them.
Zapier prices these at 3x by default. Make will not say what they cost.
Same work, different units
4 steps, run 1,000 times a month, is 6,000 Zapier tasks and 1,000 n8n executions. That is the same automation counted 6 times over.
Nothing changed except the unit. A Zapier task is one step. An n8n execution is one whole run. Any calculator that does not ask how many steps you have cannot tell these apart, which is why most of them are wrong.
| Tool | What they sell | Units this month | Cheapest plan that covers it |
|---|---|---|---|
| Zapier | tasks | 6,000 | No published priceAbove 750 on Professional. See the note below. |
| n8n Cloud | executions | 1,000 | €20/moStarter, billed annually, hosted by n8n |
| Make | credits | 4,000at minimum | $12/moCore, list price at the 10k credit tier, monthly or annual basis not verified |
| BizOps Pilotours | nothing per unit | not counted | $39/moPro, flat, monthly, cancel anytime |
Show the work: Zapier6,000 tasks
- 3 plain steps x 1 task each
- = 3 tasks
- 1 AI step x 3 tasks each (Advanced tier, Zapier's default)
- = 3 tasks
- one run of one workflow
- = 6 tasks
- 6 tasks x 1,000 runs a month
- = 6,000 tasks
This is past the 750 tasks on Professional, which is the highest tier where Zapier publishes both a price and a task allowance we could verify. The Team tier is $69/mo but its allowance is not published, and above that Zapier says the cost per task falls without saying to what. We will not guess a number here. Zapier keeps running and bills you per task as you go.
- A Zapier task is one successful action step, so steps multiply. A workflow with twice the steps costs twice as much to run once.
- Triggers, Filters and Paths do not count as tasks, so they are not in the step count above.
- Connect your own AI account and those AI steps drop from 3x to 1x: 4,000 tasks a month instead of 6,000 tasks. Cheaper, and still counted. Bringing your own key removes the markup, not the meter.
- Zapier pauses any step that reaches 75 tasks in a single run and asks you to approve it. That circuit breaker exists because AI steps run away with allowances.
- If you drive Zapier from an AI assistant instead, every successful MCP tool call costs 2 tasks out of this same allowance. That is not in the total above, because it depends on what you ask.
Read off
- https://zapier.com/pricing read 2026-07-16
- https://zapier.com/blog/zapier-pricing/ read 2026-07-16, page dated 2025-10-10
- https://help.zapier.com/hc/en-us/articles/46597632373389-AI-by-Zapier-new-model-based-pricing-starting-June-15-2026 read 2026-07-16, page dated 2026-06-15
- https://help.zapier.com/hc/en-us/articles/45645738385805-How-Zapier-MCP-usage-works read 2026-07-16, page dated 2026-06-04
Show the work: n8n Cloud1,000 executions
- 4 steps in the workflow
- = does not change the count
- 5 workflows x 200 runs each
- = 1,000 executions
- One execution is one whole workflow run, however many steps it has. This is the single biggest difference in the table, and it is why step count has to be an input.
- n8n does not count sub-workflow runs, manual runs from the editor, error-workflow runs, or polls that returned no data. Of the three metered vendors here, n8n counts the most generously.
- Prices are the annual-billing figures and are in euros, as n8n publishes them. We do not convert to dollars, because we have no sourced exchange rate and inventing one is the exact mistake this page exists to fix.
- n8n's AI Assistant has its own separate credit allowance per cloud plan. It is a builder helper, not a workflow step, so it is not modelled here.
n8n Community Edition is free, self-hosted, and unlimited, and on self-hosted you bring your own API key with no credit limits at all. If you are willing to run a server, n8n costs less than we do and gives you the same no-meter deal. That is not a trick, and we are not going to pretend otherwise. What you buy from us instead is that there is no server, no upgrades, no Docker, and no pager at 3am.
Read off
- https://n8n.io/pricing/ read 2026-07-16
- https://docs.n8n.io/build/understand-workflows/understand-executions read 2026-07-16, page dated 2026-06-23
- https://blog.n8n.io/build-without-limits-everything-you-need-to-know-about-n8ns-new-pricing/ read 2026-07-16, page dated 2025-08-07
- https://docs.n8n.io/choose-how-to-use-n8n#self-hosted-community-edition read 2026-07-16
Show the work: Make4,000 credits
- 4 module actions x 1 credit each
- = 4 credits
- 4 credits x 1,000 runs a month
- = 4,000 credits, at minimum
- A Make credit is one module action, so steps multiply here the same way Zapier tasks do.
- Make replaced operations with credits on 2025-08-27. The plan prices did not move. The unit did, so that different actions could cost different numbers of credits.
- This figure is a floor, not a forecast. Make prices AI actions at a fixed rate or varying by model and token usage and does not publish the multiple, and their own announcement says credit usage may increase if you use Make's AI Provider. Your real number is higher than the one above. We do not know by how much, and will not pretend to.
- Make gates building with your own connection to the Pro plan or higher, so bringing your own AI key on Make starts at $21/mo, not $12.
- Make's pricing page has a monthly and annual toggle whose state we could not verify, so treat these as list prices at the 10,000 credit tier.
Read off
- https://www.make.com/en/pricing read 2026-07-16
- https://community.make.com/t/introducing-credits-a-new-system-of-billing/89480 read 2026-07-16, page dated 2025-08-13
Show the work: BizOps Pilot
- 4,000 steps executed this month
- = not counted
- 1,000 runs this month
- = not counted
- the bill
- = $39 a month
- There is no unit here, so there is no multiplication to do. Steps do not matter. Runs do not matter. Runs are unlimited on every plan, including free.
- You bring your own AI key and pay the model provider directly with zero markup. That is what makes a flat price survivable, and it is the honest asterisk on this row: our $39 does not include what your AI provider charges you. Zapier's tasks and Make's credits do include model cost, which is exactly what their AI multipliers are for. Compare accordingly.
- Free is capped on how much runs at once, 3 Pilots and 1 System, never on how often. Keep more than 3 workflows running at the same time and that is Pro.
- Founding is $249 once for Pro for life, capped at 300 seats.
- 5 workflows is past the 3 the free beta keeps running at once, so this row prices Pro.
This is our product, so read this row with that in mind. We are pre-launch and small, we have no customer counts and no testimonials to show you, and Gumloop Pro is $37/mo, two dollars cheaper than us. The one thing we do that nobody else in this table does is stop counting.
Read off
- https://bizopspilot.com/ read 2026-07-16
Who is not in the table, and why
Two tools people ask about are missing. Both are missing for the same reason: we could not model them without inventing something.
Gumloop
Gumloop bills credits by tokens analyzed, not by steps, so steps and runs cannot produce a credit count. Their own docs say the same agent might cost 2 credits for a quick question and 200 for a deep research task. Any number we printed here would be fiction. Worth knowing anyway: Pro is $37/mo with 20,000+ credits, which is cheaper than we are, and their bring-your-own-key option gives 50% off AI model credits, which means you pay your model provider for every token and Gumloop still bills you half the meter.
- https://www.gumloop.com/pricing read 2026-07-16
- https://docs.gumloop.com/core-concepts/credits read 2026-07-16
- https://docs.gumloop.com/core-concepts/ai_models read 2026-07-16
Lindy
Lindy describes its usage tiers only as Standard, 3x and 7x, and publishes no absolute unit anywhere we could find, so there is nothing to calculate against. It has also repositioned as a per-seat personal AI assistant rather than a workflow automation tool, which makes it a different product from the rest of this table.
- https://www.lindy.ai/pricing read 2026-07-16
If a vendor does not publish a number, we leave the box empty rather than fill it. That is the whole editorial policy.
Try the tool with no meterWhere this calculator refuses to answer
A calculator is only as honest as the numbers it will not make up. These are the gaps, stated rather than filled. If a vendor starts publishing one of these, we will add it.
Zapier above 750 tasks
Professional is $19.99/mo billed annually for 750 tasks, and that is the highest tier where Zapier publishes both a price and an allowance. Team is $69/mo with no published task count. Zapier also says the cost per task falls as tiers rise, so extrapolating from Professional would overstate their bill. We could print a big scary number here. It would be fiction, and it would flatter us.
Make with AI steps in the workflow
Make says AI actions are priced at a fixed rate or varying by model and token usage, and does not publish the multiple. Their own credits announcement warns that your credit usage may increase if you use their AI Provider. So our Make number counts AI steps at one credit and is labelled a floor. Your real bill is higher. We do not know by how much.
Make above 10,000 credits
Their slider runs to 8M+ credits, but 10,000 is the only tier we could read prices off. No price, no number.
n8n in dollars
n8n publishes in euros, so we show euros. Converting would need an exchange rate, and we do not have a sourced one. Inventing an FX rate to make a comparison look cleaner is the same species of mistake this page exists to correct.
n8n and Make monthly versus annual
n8n's figures here are its annual-billing prices, which is what its page shows by default. Make's page has a monthly and annual toggle whose state we could not pin down, so treat Make as list price at the 10k tier. Both are flagged in the rows themselves.
Anyone's AI bill, including ours
Zapier's tasks and Make's credits include model cost, which is what their AI multipliers are for. Our $39 does not: you bring your own key and pay Anthropic or OpenAI directly at their price, with no markup from us. That is a real difference in what the number covers, and you should hold it in your head when reading the table.
Every number, and where it came from
All of it is the vendor's own page. No aggregators, no review sites, no "as reported by". Each was read on 2026-07-16. Pricing pages change without notice, so if you are about to spend money, click through and check us.
- https://zapier.com/pricing
Zapier bills per task. Free = $0 for 100 tasks/mo. Professional from $19.99. Task tiers run 100/mo to 2M/mo.
Read 2026-07-16. Page carries no date.
- https://zapier.com/blog/zapier-pricing/
Professional at $19.99/mo billed annually includes 750 tasks/mo. Over the limit, Zapier switches you to pay as you go. Filters, Paths and Formatter do not count toward task limits.
Read 2026-07-16. Page dated 2025-10-10.
- https://help.zapier.com/hc/en-us/articles/46597632373389-AI-by-Zapier-new-model-based-pricing-starting-June-15-2026
AI steps are priced by model tier: Standard 1x, Advanced 3x, Premium 5x. The default tier is Advanced. Connecting your own AI account is 1x. A step reaching 75 tasks in one run pauses the Zap for approval.
Read 2026-07-16. Page dated 2026-06-15.
- https://help.zapier.com/hc/en-us/articles/45645738385805-How-Zapier-MCP-usage-works
Each successful Zapier MCP tool call consumes 2 tasks, a fixed rate, drawn from the same plan allowance as Zaps.
Read 2026-07-16. Page dated 2026-06-04.
- https://n8n.io/pricing/
n8n bills per execution with unlimited workflows, steps and users. Annual billing: Starter EUR 20 / 2,500 executions, Pro EUR 50 / 10,000, Business EUR 667 / 40,000 (self-hosted).
Read 2026-07-16. Page carries no date.
- https://docs.n8n.io/build/understand-workflows/understand-executions
One execution is one whole workflow run regardless of step count. Sub-workflow, manual, error-workflow and empty-poll runs do not count.
Read 2026-07-16. Page dated 2026-06-23.
- https://blog.n8n.io/build-without-limits-everything-you-need-to-know-about-n8ns-new-pricing/
Unlimited workflows, steps and users on every plan. The Business Plan is self-hosted. Nothing changes for Community Edition.
Read 2026-07-16. Page dated 2025-08-07.
- https://docs.n8n.io/choose-how-to-use-n8n#self-hosted-community-edition
Self-hosted Community Edition. On self-hosted you bring your own API key, so no credit limits apply.
Read 2026-07-16. Page carries no date.
- https://www.make.com/en/pricing
Make bills per credit. Each module action counts as one credit. At the 10,000 credits/mo tier: Free $0 / 1,000 credits, Core $12, Pro $21, Teams $38. Credit tiers run 10k to 8M+.
Read 2026-07-16. Page carries no date.
- https://community.make.com/t/introducing-credits-a-new-system-of-billing/89480
Operations became credits on 2025-08-27 at unchanged prices. Most actions stay 1 credit per module run; advanced features, especially AI, are priced at a fixed rate or varying by model and token usage. Credit usage may increase with Make's AI Provider. Building with your own connection requires Pro or higher.
Read 2026-07-16. Page dated 2025-08-13.
- https://www.gumloop.com/pricing
Gumloop Pro is $37/mo with 20,000+ credits and unlimited seats. Free is 5,000 credits.
Read 2026-07-16. Page carries no date.
- https://docs.gumloop.com/core-concepts/credits
Gumloop credit costs are variable: the same agent might cost 2 credits for a quick question and 200 for a deep research task.
Read 2026-07-16. Page carries no date.
- https://docs.gumloop.com/core-concepts/ai_models
Gumloop BYOK gives 50% off AI model credits and requires the Pro plan or higher. You supply the key and still burn half the credits.
Read 2026-07-16. Page carries no date.
- https://www.lindy.ai/pricing
Lindy sells per seat with an undefined usage unit described only as Standard / 3x / 7x. Plus $49.99, Pro $99.99, Max $199.99.
Read 2026-07-16. Page dated 2026-05-04.
- https://bizopspilot.com/
Flat $39/mo Pro. $249 one-time Founding, capped at 300 seats. Free beta tier. Unlimited runs on every plan. Bring your own AI key and pay the model provider directly, zero markup.
Read 2026-07-16. Page carries no date. This one is ours.
Or stop counting altogether
Every row above except ours asks you to forecast your own bill. We think that is a spreadsheet you should not have to build. Flat price, unlimited runs, bring your own AI key and pay the model provider directly. Free while we are in beta.
Start free