Which Make Pricing Plan Fits a Small Business? Credits Explained

Coding Liquids tutorial cover featuring Sagnik Bhattacharya for Which Make Pricing Plan Fits a Small Business? Credits Explained.
Coding Liquids tutorial cover featuring Sagnik Bhattacharya for Which Make Pricing Plan Fits a Small Business? Credits Explained.

A small business running three or four AI scenarios usually fits Make's lowest paid tier of about 10,000 credits a month, because one run of a typical scenario uses 4 to 15 credits. Count credits as runs a month times modules per run, add AI modules at their own rates, and choose Teams only if you need roles.

The Complete Excel Course with AI Integration

From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.

By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner

Buy now

The error most buyers make is pricing by plan name. Core, Pro and Teams differ in features, and the credit allowance is chosen separately within each plan, so work out your credits first and your features second. Make's help centre says Core goes up to 300,000 credits a month and larger tiers sit on Pro, far more than a few small-business scenarios use. This tutorial does the credit sums for three scenarios and then shows what the features change.

How many credits does a Make scenario run use?

For ordinary apps, each module that runs uses one credit, so a month's credits are roughly runs times modules per run, plus whatever extra the AI modules use at their own rates.

Make's help page says credits replaced operations as the term for its billing unit. For ordinary apps, one operation is one credit, and fixed-rate modules cost the same regardless of the size of the data: uploading a file to Google Drive uses 1 credit per upload. Only features triggered by scenario runs, plus the chat of an AI agent, use credits.

AI changes the picture. The same page describes four cases (read 17 November 2026):

  • Third-party AI apps such as OpenAI, Anthropic Claude and Gemini: one operation is one credit, and modules with automatic AI provider connections also include token usage.
  • Make's AI Provider, in Make AI Agents and the Make AI Toolkit: credits depend on tokens and operations. The rates depend on the model and change, so Make's credits help page is the authority and no figure is quoted here.
  • A custom AI provider connection on paid plans: you pay Make credits per operation and pay your provider directly for tokens.
  • Automatic connections for Make AI Web Search, Make AI Content Extractor and the Simple Text Prompt module in certain apps: credits based on tokens, operations and other usage factors. Some Content Extractor modules use 2 or 10 credits per operation.

The upshot is that an AI step can cost more than one credit, so a scenario's credit price is the sum of its modules, not its length. For a first read on the topic, see credits explained for a booking flow.

What the plan prices are, and where to confirm them

At the time of writing, Make's pricing page lists Free, Core, Pro, Teams and Enterprise, and Make's extra-credits help page uses a Core-style plan of 10,000 credits at $9 a month as its worked example. Paid prices depend on the credit tier you pick and on whether you pay monthly or annually, and annual billing is cheaper, so the figure you pay may differ. Open Make's pricing page, set the credit slider to the tier you need and switch between monthly and annual billing before you decide.

What the page does confirm:

  • The Free plan gives 1,000 credits a month.
  • Credits expire at the end of the term; unused credits do not roll over.
  • Extra credits are bought manually in steps of 1,000 or automatically in 10,000s, at your plan's per-credit rate plus a 25 per cent surcharge, on paid plans only.
  • Extra credits expire at the end of the billing cycle on monthly billing and at the end of the year on annual billing (on annual Core, at each monthly reset).
  • Annual payment in advance costs less than monthly.
  • Companies under 20 employees may qualify for a Start-up Plan at 50 per cent off Business.
  • If you downgrade, active scenarios that use higher-plan features are stopped, and unused credits move to the new subscription.

Main worked example: three AI scenarios for a small manufacturer

Say a manufacturer with fourteen staff wants three scenarios. The credit counts below are illustrative modelling choices, to be replaced by readings from your own scenario history.

Scenario A: sort new quote enquiries

A form submission arrives (1 credit for the trigger), an AI module classifies it as standard, custom or spam (1 credit through a third-party app, plus token use if the connection is automatic), a Google Sheets row is added (1) and a Slack message is sent (1). That is 4 credits a run. With 150 enquiries a month, the scenario uses 600 credits.

Scenario B: read supplier invoices

A PDF arrives by email (1), an AI Content Extractor module reads it (2 or 10 credits depending on the module, so price both), a Sheets row is added (1) and an approval email is sent (1). That is 5 credits a run at the low end and 13 at the high end. With 120 invoices a month, the range is 600 to 1,560 credits.

Scenario C: a daily digest

Once a day a schedule fires (1), a Sheets search returns the day's rows (1), an aggregator combines them (1), an AI module writes a summary (1) and an email goes to the owner (1). That is 5 credits a run and 150 credits over 30 days. One caution: modules after a search run once per bundle returned unless an aggregator joins them, so a search that returns 40 rows can multiply downstream steps. Check this in your first test run.

ScenarioCredits a runRuns a monthCredits a month
A: enquiry sorting4150600
B: invoice reading (low / high)5 / 13120600 / 1,560
C: daily digest530150
Total (low / high)1,350 / 2,310

The answer to the headline question is now clear. The three scenarios need 1,350 to 2,310 credits a month, which is more than the Free plan's 1,000 and well inside a 10,000-credit paid tier. The manufacturer has four to seven times the headroom it needs, which is room to add scenarios, retries and test runs. A 10,000-credit tier is the sensible first purchase. Going to 20,000 would be paying for space nobody has used.

Three more credit sums

A wholesale distributor with a stock alert

A scenario checks a stock sheet each hour. If it fires 24 times a day for 30 days, that is 720 runs. If each run uses 2 credits when nothing is low, it burns 1,440 credits a month doing nothing. Slowing the check to every four hours cuts it to 180 runs and 360 credits. The tutorial on running scenarios less often goes into this. Make's credit pages do not spell out how an empty polling check is counted, so open your scenario history after a day of runs and read it.

A B2B software reseller with an AI email reply

A scenario watches a shared inbox (1), asks an AI module to draft a reply (1 credit plus tokens) and saves it as a draft (1). That is 3 credits a run. With 400 emails a month, it is 1,200 credits. If the team uses its own provider key, Make charges credits per operation and the provider bills tokens separately, so total cost sits in two places. Add the provider's token bill before you compare plans.

An HR consultancy with a weekly report

A scenario runs once a week (1), reads a tracker sheet (1), has an AI module write a summary (1) and emails four partners (4, if the email module runs once per recipient). That is 7 credits a run and about 30 a month. It would fit the Free plan on its own. Combine it with other jobs before you decide you need paid.

What does the Teams plan change for a small business?

The Teams plan lets an organisation run more than one team and manage team roles such as Team Operator and Team Monitoring; on plans below it, you get one team and cannot manage roles within it.

Make organises people inside an organisation, which holds the plan and billing, and teams within it. Make's help pages list these organisation roles: Owner, Admin, Member, Accountant, App Developer and Guest. Team roles are Team Admin, Team Member, Team Operator, Team Monitoring, Team Guest and a Team Observer that Make assigns itself.

The point that affects the plan choice is this line in the help page: on plans below the Teams pricing plan, your organisation can have only one team, and you cannot manage roles within it. A five-person firm sharing one team can run happily without roles, with everyone a Member. A firm that wants an assistant to operate scenarios (turn them on and off) without being able to edit them needs the Team Operator role, so it needs the Teams plan.

I did not find a per-seat price in the pages I read. If Make's pricing card for your plan lists users, count your headcount against it. Custom roles are an Enterprise feature, so do not plan around them.

How do you buy extra Make credits without surprises?

Top up by hand in steps of 1,000 credits rather than relying on auto-purchase, which buys 10,000 at a time, and remember that every extra credit costs 25 per cent more than your plan's own rate.

When credits run out before the next billing cycle, Make says you can upgrade, buy extra credits or switch on auto-purchase. Make's extra-credits page says extra credits cost your plan's per-credit rate plus a 25 per cent surcharge, so on its own example of 10,000 credits for $9, 1,000 extra credits cost $1.125 and 10,000 cost $11.25. Manual purchases move in steps of 1,000 credits, and they expire at the end of the billing cycle on monthly billing.

That makes auto-purchase of 10,000 credits a poor safety net for a small firm. An accidental loop can burn a 10,000 bundle quickly, and the purchase happens without your involvement. Make caps auto-purchases at your plan's credit amount per billing cycle, but that is still a lot of unplanned spend. Better to set an internal alert at 70 per cent of monthly credits, and top up by hand in steps of 1,000. Check the current surcharge on Make's extra-credits help page before relying on this.

What mistake can double a Make bill?

Testing a new scenario on a whole folder of real files, especially when a bug sends each file through the AI step more than once, can use more credits in one afternoon than the Free plan holds.

Take a distributor that builds an invoice scenario and tests it with a folder of 60 old PDFs. Each test run uses credits, and a bug in the filter sends every PDF through the AI step three times. At 13 credits a run that is 60 files times 3 runs times 13 credits: 2,340 credits in one afternoon, more than the Free plan holds. The lesson is to test with five files, to read the credit use after each test and to set the scenario's schedule to on demand until it behaves.

A planning sheet to copy

Make credit planner (one row per scenario)

Scenario: ______________  Runs a month: ____
Modules per run (count each one that fires):
  trigger ... __  app actions ... __  AI module(s) ... __ (use its own rate)
  iterators / searches that return many rows ... __ x rows
Credits a run = ____      Credits a month = runs x credits = ____

Total across scenarios: ____
Add 30 per cent for tests and retries: ____
Pick the smallest tier at or above that number.

Filled in (illustrative):
A enquiry sorting 150 x 4 = 600 | B invoices 120 x 13 = 1,560
C digest 30 x 5 = 150          | Total 2,310 | +30% = 3,003
Tier: 10,000 credits. Teams only if roles are needed.

If you would rather keep this planner as a spreadsheet, try The Complete Excel Course with AI Integration on Coding Liquids.

How do you cut a scenario's credits?

Put the cheapest check first and the AI step last: a filter straight after the trigger stops items that do not matter, and modules that never run use no credits.

Scenario B is the expensive one, so it is the place to save. Suppose the shared inbox receives 120 emails a month, but only 72 of them are real invoices; the rest are statements, newsletters and replies. As first built, every email passes through the trigger, the extractor, the Sheets step and the approval email, at the high rate of 13 credits: 120 times 13 is 1,560 credits.

Now add a filter straight after the trigger that lets through only messages with a PDF attachment and "invoice" in the subject or file name. The trigger still counts for all 120 emails (120 credits), but the later steps, about 12 credits at the high rate, run only for the 72 invoices: 72 times 12 is 864. The new total is 984 credits, a saving of 576 a month for one filter. Modules that never run do not use credits, which is the whole logic behind filters and routers. Modules that do not run on a given run are not charged, so only the branch that runs costs credits. Read one router run in your scenario history to see whether the router itself shows a credit charge before you build around it.

The same thinking applies to Scenario A. If 30 of the 150 enquiries are obviously spam, a cheap keyword filter before the AI classifier saves the AI step on those 30. The principle is to put the cheapest possible check first and the AI step last.

Why does a digest scenario cost more without an aggregator?

Without an aggregator, each later step runs once per row instead of once per scenario run, so a digest costs far more credits than its module count suggests.

Scenario C looks cheap at 5 credits a run, and it is, until the aggregator is missing. Say the Sheets search returns 40 rows. Without an aggregator, the AI summary and the email each run once per row. The run then costs 1 for the schedule, 1 for the search, 40 for the AI step and 40 for the email: 82 credits instead of 5. Over 30 days that is 2,460 credits, more than the other two scenarios combined, to send a digest that arrives as 40 separate emails. Spot it in the first test: if the run log shows 40 bundles reaching the email module, stop the scenario and add the aggregator before the AI step.

A prompt for the classifier, with a sample reply

Credits are only half the cost of an AI scenario; the prompt decides whether the output is usable. This is an illustrative prompt for Scenario A's classifier and a realistic reply, with what you would fix.

PROMPT (inside the AI module)
You sort quote enquiries for a small manufacturer. Read the message
below and reply with JSON only:
{"type": "standard" | "custom" | "spam", "urgency": "low" | "high",
 "summary": "max 20 words"}
Rules: "custom" means the customer needs a non-catalogue dimension
or material. "high" urgency means a stated deadline within 7 days.
Message: {{1.message}}

SAMPLE REPLY (illustrative)
{"type": "custom", "urgency": "high",
 "summary": "Needs 40 steel brackets, 12 mm hole, delivered by Friday."}

WHAT TO FIX
- If the reply arrives wrapped in a sentence, the next module cannot
  read it: add "no text outside the JSON".
- Test ten odd messages (blank, very long, another language) before
  going live, and add a fallback value for "type".

Each test message costs credits, so choose the ten oddest messages you have rather than a hundred ordinary ones.

How do you read your real credit use after week one?

Open each scenario's history in Make after a week of live runs, note how many runs there were and the credits a typical run used, and compare both with your planning table before you re-pick the tier.

A model is a guess until you compare it with reality. After a week of live runs:

  1. Open each scenario's history in Make and count how many executions ran.
  2. Check the credits recorded for a typical run, and compare them with your table. If a run used more than you expected, open it and look for a module that ran several times, such as one after a search.
  3. Multiply the real credits a run by your monthly volume. That number replaces your guess.
  4. Compare the total against your tier. If you are using less than a third of it, you may be able to drop a tier at the next renewal; if you are above 70 per cent in week four, plan a top-up or a higher tier.

Write the three readings (runs, credits a run, total) beside the table. The next time you add a scenario, you will price it from measured numbers, not hope.

Annual or monthly billing for a small firm

Annual billing costs less per month, and the pricing page says so, but the money is paid in advance and unused credits expire at the end of the term in any case. A small firm in its first quarter of automation should start monthly. The first three months usually show the real credit use, and a firm that discovers it needs 2,000 credits a month instead of 10,000 would have locked in the wrong tier for a year. After a quarter of stable use, annual billing becomes sensible, and the saving is real. The table below is a way to decide, with no prices needed:

Your situationBilling choice
First scenarios, volumes unknownMonthly
Three months of stable credit use, under 70 per cent of the tierAnnual, same tier
Volumes rising each monthMonthly, so you can move up
Employee count under 20 and a larger plan wantedAsk about the Start-up Plan first

When Make is the wrong thing to pay for

If your three scenarios amount to under 1,000 credits a month, the Free plan covers you, and the only reason to pay is a feature. If a scenario is one email each time a form comes in, your form tool may do it natively. Compare against Zapier and n8n before you commit, because the lowest bill depends on how many steps your scenarios have. To see which scenarios are worth building at all, five scenarios a small firm can run without code is a good start.

Make restructured its plans in 2026 and changes prices from time to time, so check the pricing page on the day you buy, and treat the credit sums above as a method rather than a quote.

If you want help deciding which scenarios are worth building for your business, I offer one-to-one AI consultations for small businesses.

Make pricing for small businesses: questions owners ask

Do unused Make credits roll over to next month?

No. Make's pricing page says credits expire at the end of the term, so unused credits do not roll over. Extra credits you buy expire after one month on monthly subscriptions, or at the end of the year on annual Pro and Teams subscriptions. Buy the tier you will use, not the tier you hope to grow into.

Does every Make module cost one credit?

For ordinary apps, one operation uses one credit. AI modules differ. Calls through third-party AI apps count one credit per operation, while Make's own AI Provider and some automatic AI connections add token-based usage. Some Make AI Content Extractor modules use 2 or 10 credits per operation, so check each AI module before you price a scenario.

Can a small team manage roles on a cheaper Make plan?

Not fully. Make's help page says that on plans below the Teams pricing plan an organisation can have only one team and you cannot manage roles within it. Organisation roles such as Owner, Admin, Member and Accountant still exist. If you need Team Operator or Team Monitoring roles for staff, plan for the Teams tier.

Is there a discount for small companies on Make?

Possibly. Make's pricing page says companies under 20 employees may qualify for its Start-up Plan, described as 50 per cent off Business. Eligibility is decided by Make, so read the current terms and ask before you pay full price. Annual billing is also cheaper than monthly.

Further reads

Sources: Make's pricing page, its credits help page and its organisation and team help pages, read on 17 November 2026.

Rather not work this out alone?

An AI implementation consultation is a one-to-one call with me, not an agency or a sales team. You leave knowing what AI you can use in your business and where to start.

Book a 1:1 call with me