Make credits explained: what one operation costs in six real automations
Short answer
In Make, one operation is one module run, and for most modules one operation costs one credit; a trigger check costs one whether or not it finds anything, and a module that receives ten bundles runs ten times. So a three-module webhook automation costs three credits per event, a daily search costs one per day plus the modules per row found, and a Gmail watch polled every 15 minutes costs about 2,976 a month before it does any work. The six worked examples below show the arithmetic.
Credit, operation, scenario, bundle, polling, webhook, interval, blueprint: each defined once from Make's docs, then costed in six real automations.
Disclosure: links marked "paid link" pay us a commission from the vendor if you sign up. Your price is unchanged, and every file here works without them. How we choose tools →
Every Make pricing conversation ends in the same place: "so what does one credit actually buy?" The definitions are short, they come from Make's own help centre, and once they are clear, the monthly cost of any automation is arithmetic. This page gives the definitions once and then does the arithmetic for six of our blueprints. Other guides on this site link here instead of repeating it.
Definitions are quoted or closely paraphrased from Make's Help Center and pricing page as read on 29 September 2026, linked at the bottom. The credit figures for our automations are computed from the exported blueprint files with the same code that runs our calculator, at the calculator's default volumes and a 31-day month.
The terms, defined once
| Term | What it means | Why it matters for the bill |
|---|---|---|
| Credit | The currency you buy and consume to use Make. Plans include a monthly number: 1,000 on Free, 10,000 on Core at $12/month (checked 2026-09-29). | It is the unit on the invoice. For most features one operation equals one credit at a fixed rate, regardless of data size. |
| Operation | A single module run to process data or check for new data. | One operation per module per bundle; one per trigger check. Count operations and you have counted credits for standard modules. |
| Scenario | One automation: a chain of modules starting with a trigger. Free runs 2 active at a time; Core has no cap. | The scenario cap, not credits, is usually what ends the Free plan. |
| Module | One step in a scenario: a trigger, an action (create, update, send), a search, or a tool such as a router, filter or aggregator. | Every module that runs uses an operation for each bundle it processes. |
| Bundle | One item of data flowing through the scenario: one lead, one row, one email. | Each bundle triggers its own module run. A search that returns ten rows makes every module after it run ten times. |
| Trigger | The first module. It decides when the scenario runs. | Instant (webhook) triggers cost nothing while idle. Polling triggers cost one operation per check, found or not. |
| Webhook (instant trigger) | A URL Make creates; an external app calls it when something happens, and the scenario runs immediately. | Zero operations while nothing happens. The cheapest and fastest trigger. |
| Polling (scheduled "watch" trigger) | A trigger that asks an app for new data on a schedule. | One operation per check. At 15 minutes: 96 a day, about 2,976 in a 31-day month. |
| Interval | How often a scheduled scenario runs. The minimum depends on your plan: 15 minutes on Free, 1 minute on Core. | Interval × polling trigger = the idle cost of the scenario. |
| Data transfer | The volume of data your scenarios move: 512 MB a month on Free, 5 GB on Core. | Only matters when moving files. Leads and rows are kilobytes. |
| Blueprint | A reusable version of a scenario: its modules, module settings and mapped values, exported as a JSON file. | Connections and credentials are not included; you add your own after import. It is how our automations are shared. |
The three counting rules
- One module run is one operation, and for standard modules one operation is one credit. A scenario with three modules that runs once uses three credits.
- A trigger check is one operation whether or not it finds anything. Make's Operations page: trigger modules run once to check for or retrieve data, regardless of the number of bundles returned. That is what makes polling expensive and webhooks cheap.
- Bundles multiply. Every module after a search or a watch runs once per bundle. Gmail "Send an Email" fed five bundles sends five emails and uses five operations. A search that finds twenty rows makes the two modules after it run twenty times each: 1 + 20 + 20 = 41 operations in one run.
Two more things to know. AI modules can use more than one credit per operation, priced on tokens, so an automation that calls a language model is not "one credit per step". And extra credits, bought mid-month on a paid plan, cost 25% more than plan credits; the Free plan cannot buy them at all (checked 2026-09-29).
Six automations, costed
Each row below is one of our blueprints. "Modules" is read from the exported file. "Runs a month" is the calculator's default volume for that automation, or the number of scheduled runs in a 31-day month: 31 daily, 5 weekly, 1 monthly. The credits column is the result of the three rules above, and it matches what the calculator shows for the same inputs.
| Automation | Trigger type | Modules | Runs or items a month | The arithmetic | Credits a month |
|---|---|---|---|---|---|
| Lead alert: form → Sheets row → Telegram | Webhook | 3 | 100 leads | 100 × 3 | 300 |
| Lead duplicate check: form → search sheet → row → alert | Webhook | 4 | 100 leads | 100 × 4 | 400 |
| Invoice reminder: daily search → email → mark sent | Daily search | 3 | 31 runs · 20 reminders | 31 + 20 × 2 | 71 |
| No-show reminder: daily search → email → mark sent | Daily search | 3 | 31 runs · 40 reminders | 31 + 40 × 2 | 111 |
| Weekly report: Sunday search → aggregate → Telegram | Weekly summary | 5 | 5 runs | 5 × 5 | 25 |
| Email → task: watch Gmail label → row → Telegram | Polling Gmail | 3 | 2,976 checks at 15 min · 40 emails | 2,976 + 40 × 2 | 3,056 (824 hourly · 111 daily) |
Reading the table
The first two rows are webhook scenarios: cost is volume times modules, and nothing while idle. Double the leads and the bill doubles; no leads, no credits. The lead alert at 100 leads is 300 credits; at 300 leads it would be 900, still inside Make's 1,000-credit Free plan on credits alone (checked 2026-09-29).
Rows three and four are daily searches. The search runs every morning: 31 operations a month regardless. Then the email and the update run once per row found. Twenty overdue invoices in a month is 40 more operations; 71 in total. This is the shape to prefer for anything date-based: a fixed, small daily cost plus a cost proportional to the work actually done.
Row five is a summary scenario. The search returns many rows, but an aggregator folds them into one bundle before the message module, so every module runs once per run: five modules, five Sundays, 25 credits. Aggregators are how you stop bundles multiplying.
Row six is the warning. The same three-module shape as the lead alert, but the trigger polls Gmail instead of waiting for a webhook. At the Free plan's 15-minute floor (checked 2026-09-29) the trigger alone is 2,976 operations a month, three times the Free allowance, before the 40 emails add their 80. Poll hourly and the total is 824; poll once a day and it is 111. Nothing in the scenario changed except the interval.
The blueprints in the table
Lead comes in → logged → alert. No hands.
A form on the client site sends a lead; it is logged and an alert reaches Telegram in seconds. The first automation every consultant sets up.
Blueprint, steps and caveats → · Watch the tutorial on YouTube
One new automation a week, by email
Returning lead → recognised → "second contact", not "new lead"
A returning lead is greeted as new. Before logging, the phone is checked against the log; if it is there, the alert says "came back", not "new lead".
Invoice past due → an automatic, polite reminder
Once a day it finds open invoices past their date and sends a reminder. The client no longer has to remember to chase.
A meeting tomorrow → a reminder 24 hours ahead, with a "can't make it" link
A no-show is a slot nobody could refill. Every morning, each of tomorrow's meetings gets a reminder email with the time and a "can't make it" link.
Weekly report → Telegram, every Sunday at eight
The numbers the client asks about every week — leads, open tasks, invoices — reach them on their own, in one message, before they ask.
Labelled email → task in the log → reminder
Tag an email with one label and it becomes a task with a due date and an alert. One screen for everything that is waiting.
How to cost your own scenario
- Count the modules, including the trigger and anything inside router routes. Filters between modules are not modules.
- Name the trigger type. Webhook: skip to step 4. Scheduled search: note the schedule. Polling watch: note the interval.
- Compute the fixed cost. Scheduled search: runs a month (31 daily, 5 weekly, 1 monthly). Polling: 31 × 1,440 ÷ interval in minutes, rounded up (15 min → 2,976; 60 → 744; 1,440 → 31).
- Compute the per-item cost. Webhook: events × modules. Search or polling: items × (modules − 1). Summary with an aggregator: runs × modules, and no per-item cost.
- Add 20% for test runs, re-runs after an error and busy months. Compare the total with the plan: 1,000 on Free, 10,000 on Core (checked 2026-09-29). Then check the scenario count against the Free plan's cap of two.
Or let the calculator do it (free)Pick the automations, type your volumes; the same rules, the same blueprint files
Every number on this page was measured in Make; the free plan is enough to import a blueprint and watch the operations counter yourself: Sign up (one month of Core free) → paid link · we earn a commissionPaid link. If you sign up through it we earn a commission from Make; your price is the same, and the link gives you one month of Core free.
What does not work
- Estimating from the number of "steps" a business owner describes. The blueprint's module count is what Make bills. Read the file.
- Forgetting the trigger. On a polling scenario the trigger is usually most of the cost. On a webhook scenario it is zero. The trigger type matters more than the module count.
- Assuming one credit per operation for AI modules. Make prices them on tokens as well as operations, and the rate differs by model (checked 2026-09-29).
- Comparing Make credits with Zapier tasks one to one. Zapier does not count triggers or its built-in filters and formatters; Make counts every module run. Price the same workflow in each tool's own unit.
- Reading our default volumes as your bill. A hundred leads a month is the calculator's starting point, not a forecast for your business.
What is a Make credit?
The currency you buy and consume to use Make. Plans include a monthly allowance, 1,000 on Free and 10,000 on Core at $12/month, and for most features one operation costs one credit at a fixed rate regardless of data size (Make Help Center and pricing page, checked 2026-09-29).
What is an operation in Make?
A single module run to process data or check for new data. A module fed five bundles runs five times and uses five operations; a trigger check uses one whether or not it finds anything.
Do webhooks use credits?
A webhook trigger uses an operation when it receives data and runs the scenario, and nothing while waiting. That is why a webhook scenario's monthly cost is simply events times modules.
How many credits does polling every 15 minutes use?
About 2,976 a month for the trigger alone in a 31-day month (96 checks a day), before the rest of the scenario runs on anything found. Hourly is 744; once a day is 31.
What is the difference between a scenario and a blueprint?
A scenario is the live automation in your account, with its connections. A blueprint is the exported JSON of its modules, settings and mappings, without connections or credentials; you import it and add your own accounts.
Does an aggregator save credits?
Yes, when many rows should become one message. An aggregator folds the bundles into one, so the modules after it run once per run instead of once per row. Our weekly report costs 25 credits a month for that reason.
- Make Help Center — Credits ("Credits are the currency you buy and consume to use Make"; one operation equals one credit for most features; AI modules priced on tokens)
- Make Help Center — Operations ("An operation is a single module run to process data or check for new data"; trigger modules run once per check; bundles multiply)
- Make Help Center — Scenario blueprints (modules, settings and mapped values; connections created after import)
- Make Help Center — Webhooks (a URL called from an external app to trigger a scenario; queue and rate limits)
- Make Help Center — Schedule a scenario (interval options; minimum depends on plan)
- Make Help Center — Scenario execution, cycles and phases
- Make Help Center — Extra credits (25% additional cost; paid plans only)
- Make — Pricing (Free 1,000 credits, 2 active scenarios, 15-minute interval, 512 MB; Core $12/month, 10,000 credits, 1 minute, 5 GB; checked 2026-09-29)
Related guides
Next step
All automations →
BOOKING LOGBooking form → meetings sheet → confirmation email. Feeds the no-show reminder
A booking arrives and nobody confirms it. A Tally form writes the meeting to the sheet, emails the client a confirmation and pings you on Telegram.
AUTO WELCOMENew client → welcome email → onboarding checklist → alert
A client signs and hears nothing until someone is free. On signing, they get an email saying what comes next, a checklist opens and the team is alerted.
ONE QUESTIONFeedback score → the same thank-you and review link for everyone; a low score also alerts you
Unhappy clients stay silent until the review. A one-question Tally form (0–10) sends everyone the same review link; a 6 or lower also pings Telegram.
Get one new Make automation each week
Get the full 21-blueprint pack (ZIP) by email, then one new automation a week. Everything on the site stays free to download without it.