Make.com Pricing in 2026: What the Credits Actually Cost

How Make.com pricing actually works

Make.com prices on credits, and a credit is consumed by nearly every module that runs in a workflow ("scenario"), not just the trigger that kicks it off. Make's own documentation explains the mechanic plainly: an operation is a single module run to process or check data, and most modules consume one credit per operation, with some AI-heavy features consuming more. A trigger checking for new data costs a credit. Every action after it, adding a spreadsheet row, sending an email, calling an API, costs another. Loop over 10 items and the module downstream runs 10 times, for 10 more credits.

That compounding is the part that surprises people coming from a flat per-seat tool. A five-step scenario running 100 times a month uses roughly 500 credits. A 20-step scenario running 500 times a month can burn through 10,000 credits, Make's entire baseline plan allowance, from one workflow alone.

Make.com's current pricing tiers

Pulled directly from make.com/en/pricing as of September 2026, the published tiers are:

Every paid tier is priced against the same credit-volume ladder (10k, 20k, 40k, 80k, up to 8M+ credits/month), so cost rises with volume regardless of which feature tier is picked. Make advertises a savings of 15% or more for paying annually instead of monthly, though the exact annual number depends on which credit volume is selected. The Free plan has no time limit, but runs on a 15-minute minimum interval between scenario executions, a real constraint for anything time-sensitive.

The published price is the annual price

One detail worth knowing before you budget: the headline numbers above are the annual-billing prices. Make's pricing page defaults to the annual toggle, and the monthly figures are meaningfully higher. The full ladder below was read directly off make.com/en/pricing on September 18, 2026, and it is the single most useful table on this page because it shows both axes at once: the feature tier you pick, and the credit volume you commit to.

Two patterns fall out of it. First, monthly billing costs roughly a third more than annual at every step, so the $12 Core plan most buyers quote is really $16 unless they commit for a year. Second, the per-credit price falls as volume rises: Core at 10,000 credits works out to $0.0012 per credit on annual billing, while Core at 300,000 works out to about $0.0008. Make says this itself in its pricing-change notes, recommending that anyone regularly buying extra credits move up a tier because higher tiers have lower per-credit costs.

Core also has a ceiling. It tops out at 300,000 credits per month, and volumes above that sit on Pro or higher.

Make's own November 2025 pricing update also changed how overage works: extra credits, whether purchased manually or auto-purchased when a plan's limit is hit, now cost a flat 25% more than the credits included in the plan. Previously, manual purchases carried no markup and only auto-purchases did. Anyone running close to their monthly limit is now paying that premium either way.

What actually drives the cost up

The credit model rewards knowing exactly what a workflow will do before building it, which is harder than it sounds for a first scenario. Two things push real bills higher than the sticker price on the plan page suggests:

Polling triggers that check on a schedule burn credits on every check, whether or not there's new data. A trigger set to poll every minute runs 1,440 times a day, consuming credits even during periods when nothing changed. Webhook-based triggers, which fire only when an event actually happens, avoid this, but not every app supports a webhook.

AI-related modules and the Make Code app can consume more than one credit per run, sometimes calculated per second of execution time rather than per action. A scenario that looks identical to a basic data-sync workflow can cost meaningfully more once an AI step or a longer-running code block is added.

Worked scenarios, with the assumptions on the table

None of what follows is survey data. It is arithmetic from Make's own credit rules, with every assumption written down so you can substitute your own numbers.

A small workflow that fits in the free tier. Assume one scenario with five modules, triggered 100 times a month. That is roughly 500 credits, comfortably inside the Free plan's 1,000. The catch is not credits, it is timing: Free enforces a 15-minute minimum interval between scheduled runs, caps you at 2 active scenarios, limits a single execution to 5 minutes and files to 5 MB, and keeps execution logs for 7 days. For a personal automation that is fine. For anything a customer notices, the 15-minute floor is the thing that pushes you to Core, not the credit count.

A mid-sized operations workload. Assume one 12-module scenario running 1,500 times a month, which is 18,000 credits, plus a second 6-module scenario running 500 times, another 3,000. Total 21,000 credits a month. That lands on the 40,000-credit step rather than the 20,000 one, because credits do not roll over on monthly billing and running out stops your scenarios. Core at 40,000 credits is $38/month annual or $51 monthly. Note what just happened: the plan you need is decided by the volume ladder, not by the feature tier, and most buyers shop the feature tier first.

The polling trap, priced. This is the scenario that generates the surprise invoices. Assume one trigger polling every minute. That is 1,440 checks a day, roughly 43,200 credits a month, before a single downstream action runs, and whether or not any new data existed. That one trigger alone overruns the 40,000-credit step. Move the same trigger to a 15-minute schedule and it costs about 2,880 credits a month, a 15x reduction for a configuration change. Move it to a webhook, where supported, and the trigger only consumes credits when something actually happened.

What the same work costs on the competitors. Take the mid-sized case above, 21,000 module runs a month. On Make that is roughly $38/month annual on Core. Zapier meters completed tasks rather than module runs, with Professional starting at $19.99/month for 750 tasks, so 21,000 task-equivalents sits far up its volume ladder. n8n meters full workflow executions regardless of step count, so the same 2,000 workflow runs would fit inside its Pro plan's 10,000 executions at €50/month on annual billing, per n8n's pricing page. That is the shape of the tradeoff in one example: Make's unit is small and cheap and multiplies with step count, n8n's unit is large and indifferent to step count, and Zapier's unit sits in between at the highest price per unit.

How to cut a Make bill

Every lever below maps to something documented in Make's own help center or pricing page, so each one is checkable.

Stop polling where a webhook exists. As the arithmetic above shows, this is worth more than every other lever combined for scheduled scenarios.

Use routers and error handlers, which are free. Make's pricing FAQ lists the exceptions explicitly: the Router module consumes no credits, and neither do the error-handler modules Rollback, Break, Resume, Commit and Ignore. Structuring a scenario so a router splits work early, rather than running modules and discarding results downstream, moves work into the parts of the canvas that do not bill.

Filter as early as possible. Every module downstream of a filter only runs on the records that pass it. A filter placed after three enrichment steps pays for enrichment on records you were going to drop.

Watch the Code app. Make's own plan comparison prices the Make Code App at 2 credits per second of code execution time, on every tier. A code block that takes 8 seconds costs 16 credits, not 1. AI modules are similarly metered dynamically rather than at a flat 1 credit.

Right-size the plan rather than buying extra credits. Extra credits cost 25% more than the credits included in your plan, whether bought manually or auto-purchased. If you buy extras most months, the cheaper move is the next volume step, where the per-credit rate is also lower.

Use the 75% and 90% notifications as a planning signal. Make notifies you as you approach 75% and again at 90% of your purchased credits. Treat the first 75% alert that arrives early in a month as the trigger to re-examine a scenario, not as a prompt to buy a bundle.

Consider annual billing for credit flexibility, not just the discount. On annual billing, prepaid credits expire after 12 months rather than at the end of each month, which smooths out seasonal workloads. Credits still expire at the end of the term either way.

Check your usage allowance, not just your credits. Data transfer, data storage, incomplete-execution storage and webhook queue size all scale with credits purchased: 5 GB of data transfer, 10 MB of data storage and a 667-item webhook queue per 10,000 credits per month. A file-heavy workflow can hit the transfer ceiling well before the credit ceiling.

What buyers and reviewers actually say about it

Zapier's own comparison of Make.com pricing lays out the credit mechanic in detail and is worth reading even though Zapier is a direct competitor: it correctly describes a five-step scenario running 100 times a month as a 500-credit workflow, and flags that some actions, like Make's Code app, cost more than one credit per execution. Zapier's separate value comparison piece argues that its own per-task billing is more predictable because Zapier only charges for completed work actions, while Make bills for steps that fire regardless of outcome, including polling checks and, depending on configuration, some error and test runs.

That's a fair point about predictability, and it's also exactly the argument you'd expect the competitor being compared against to make. The credit model is genuinely cheaper per unit of work for most mid-volume, multi-step automations, which is why independent reviewers consistently rank Make ahead of Zapier on raw cost per operation. The tradeoff is that estimating total monthly credits before building a scenario takes more upfront modeling than a flat per-task or per-seat number does.

How Make.com compares to n8n and Zapier

All three of the major workflow tools price on a different unit, which makes direct comparison harder than it looks:

n8n's unit is the most forgiving of the three for complex workflows: n8n's own pricing page confirms that a full workflow run counts as one execution "with unlimited steps," so a 20-step n8n workflow costs the same as a 2-step one. That makes n8n cheaper at high step-count, low-run-frequency volumes, and Make cheaper at high-run-frequency, moderate-step-count volumes, which is why the "cheapest" answer genuinely depends on the shape of the workflow being priced, not just its size. n8n's cloud plans also require more technical comfort to configure well; it's built for teams willing to work closer to the node level, where Make's visual canvas is the more approachable option for non-engineers. A closer breakdown of n8n specifically, including when the self-hosted option changes the math, is in Coworker's n8n alternatives guide.

Zapier's per-task model is the simplest to estimate before building anything, since a task is defined the same way regardless of how the automation is built, but it's also the most expensive per unit of work for anything beyond a simple two-step zap. The category-wide pattern, where none of the three vendors uses the same billing unit, is also common outside pure workflow tools: quote-based enterprise platforms show the same lack of a single comparable number, as covered in Coworker's breakdown of Workato's pricing.

Where Workato fits, and why it has no number

Add a fourth option and the comparison stops being a table at all. Workato's pricing page, checked September 18, 2026, publishes no prices: the page is a positioning statement and a "schedule a demo" button. That is not an oversight, it is the enterprise iPaaS model: price depends on a bundle negotiated per customer, so there is no published number to compare against.

The practical consequence for a buyer comparing all four: three of them can be modelled in a spreadsheet this afternoon and the fourth cannot be modelled at all until you have taken a call. That is worth knowing before you put all four in the same evaluation matrix and wait on one quote to fill a cell. Our breakdown of what those quotes tend to contain is in Workato pricing, and the alternatives landscape is in Workato alternatives.

One more structural difference hides inside that table. Make and n8n both include unlimited users on every plan, so collaboration costs nothing extra. Zapier's Free and Professional plans are single-seat, and multi-user collaboration starts at the Team plan, currently from $69/month for 25 users, per Zapier's pricing page. For a five-person ops team, that seat difference can outweigh the per-unit comparison entirely.

What has changed in Make's pricing, and what it means today

Make's pricing has moved twice in the recent past in ways that still shape what you pay.

August 27, 2025: operations became credits. Make replaced operations with credits as its billing unit, converting existing balances at a 1:1 ratio and leaving plan prices unchanged. For simple app modules nothing moved: one module run still equals one credit. The reason for the rename is the part that matters going forward. Credits let Make meter AI features dynamically rather than at a flat rate per run, which is why AI modules and the Code app can consume more than one credit per execution while a Google Sheets module still consumes exactly one.

November 6, 2025: extra credits got more expensive for careful buyers. The same pricing update that unified overage at 25% actually cut the cost of automatic purchases, which had carried a 30% markup, while raising manual purchases, which had carried none. If you were the kind of customer who topped up by hand to avoid the auto-purchase premium, that strategy stopped working on that date. The same update raised Pro's ceiling to 8 million credits per month, capped Core at 300,000, and made custom AI provider connections available on all paid plans, so you can now point Make's AI features at your own OpenAI or Anthropic key and pay that provider directly for tokens instead of paying Make in credits.

One note on sources. Third-party pricing trackers have logged further packaging changes during 2026, including a claimed consolidation of the Core, Pro and Teams ladder into a single plan. Make's live pricing page on September 18, 2026 still shows Free, Core, Pro, Teams and Enterprise with the volume slider described above, so that is what this post uses. When a tracker and a vendor's own page disagree, the vendor's live page wins, and it is worth loading it yourself before signing anything.

Migration and switching costs

Workflow tools look portable and mostly are not. Here is what actually moves.

Within Make, blueprints move everything except the connections. Make's blueprint documentation confirms you can export a scenario as a `.json` file from the three-dot menu in the scenario builder, and import it into another scenario or another account. The blueprint carries modules, module settings and mapped values. It does not carry your account connections, so every app has to be reauthorized after import, and an imported blueprint must be under 2 MB. That makes blueprints excellent for backup, templating and moving work between Make accounts.

Between vendors, nothing moves. A Make blueprint is a Make-format file. There is no supported path that turns it into a Zapier zap or an n8n workflow, and no path in the other direction either. Switching platforms means rebuilding every scenario by hand, reauthorizing every connection, and re-testing every error path. For a team with 40 live scenarios, that is the real switching cost, and it dwarfs any monthly price difference in the first year.

Downgrading inside Make has a cost too. Make's pricing FAQ is explicit that if you downgrade, any active scenarios using features from a higher plan are stopped, and you are notified before the downgrade completes. Your card is charged immediately, your billing cycle resets from the downgrade date, and unused credits transfer to the new subscription. So dropping from Teams to Core to save money can quietly turn off the scenarios that used team-only features.

What to do about it. Export blueprints for your critical scenarios on a schedule and keep them in version control. That costs nothing, protects you from account-level accidents, and is the only artifact of your automation work that exists outside the vendor.

Billing surprises, and how to avoid each one

Scenarios stop when credits run out. They do not queue indefinitely or degrade. Incoming webhooks queue up to your available webhook queue storage and process once credits are added, and polling triggers catch up from their last successful run, but anything time-sensitive is late. The 75% and 90% notifications are your only early warning on the self-serve tiers. Overage protection, which keeps scenarios executing past the credit limit, is an Enterprise feature.

Auto-purchase is convenient and priced accordingly. Enabling it buys 10,000 credits automatically before you run out, at the same 25% premium as a manual purchase. It is available on Core, Pro and Teams.

Credits expire. Unused credits expire at the end of the term, monthly on monthly billing and annually on annual billing. Extra credits are treated the same way. There is no perpetual balance to draw down.

The displayed price is annual. Worth repeating because it is the single most common error in Make comparison posts, including ones that copy each other: Core is $12 on annual billing and $16 month to month.

AI and code steps are not one credit. The Code app bills 2 credits per second of execution, and AI modules meter dynamically. A scenario that looked like a 10-credit run during testing with a small payload can cost several times that in production with a real one.

Usage allowances scale with credits, not with your plan name. Data transfer at 5 GB per 10,000 credits per month is the one that bites file-heavy workflows, since it has nothing to do with how many modules ran.

Who Make is genuinely wrong for

Make is a strong product at a genuinely low per-unit price, and it is still the wrong choice for several kinds of buyer.

Teams running very step-heavy workflows at low frequency. This is the case n8n is priced for. A 30-module scenario running 2,000 times a month costs 60,000 credits on Make and 2,000 executions on n8n. If your automations look like that, Make's unit economics work against you and n8n's work for you. The detail is in our n8n alternatives guide and n8n pricing.

Buyers who need a fixed, forecastable line item. Credit metering rewards teams willing to model their workflows before building them. If nobody on your team will do that modelling, you will discover your real cost in month three. A flat per-seat tool is worse value per unit of work and easier to budget.

Anyone who needs guaranteed execution past a limit. Overage protection sits on Enterprise. On self-serve plans, hitting zero credits means stopped scenarios, which is not acceptable for anything customer-facing without headroom built in.

Very simple, very low-volume automations. Two apps and one trigger, a few times a day, is a case where Zapier's simpler task model is easier to reason about and the price difference is a rounding error. Make's advantage compounds with complexity, so buy it when you have complexity.

Non-technical teams without an owner. The visual canvas is genuinely approachable, more so than n8n's node-level model, but credit optimization is an engineering activity. Someone has to care about where the polling triggers are. Without that person, bills drift upward quietly.

What this means if you're comparing options

Make, Zapier, and n8n all solve the same underlying problem: moving data and triggering actions between business systems on a schedule or event. Picking between them is a real decision, but it's a different decision from the one a team should be making about AI tool access.

If the actual need is connecting an AI assistant, like Claude, Cursor, or ChatGPT, to company systems so it can look things up and take action inside them, that's not a workflow automation problem in the Make/Zapier/n8n sense. It's a context and access problem for the AI tool itself. Coworker MCP connects AI tools to 50+ apps through one governed connection, at Coworker's own published pricing rather than a credit or task meter. It's a different layer than Make: Make automates workflows between systems on its own schedule, while Coworker MCP gives an AI tool the access it needs to act inside those systems when a person or an agent asks it to. A team evaluating both isn't choosing one instead of the other; they solve adjacent problems and often sit in the same stack. For a broader view of where workflow automation tools fit against AI-native approaches, see Coworker's guide to enterprise AI workflow automation.

Book a demo to see how Coworker MCP's pricing works, no credit calculator required.

Frequently asked questions

How much does Make.com cost per month?

Make.com's published tiers, as of September 2026, run Free ($0, 1,000 credits/month), Core ($12/month), Pro ($21/month), and Teams ($38/month), each based around a 10,000-credit monthly allowance that scales to higher volumes. Enterprise pricing is custom.

What is a credit in Make.com?

A credit is consumed by most module runs in a scenario, called an operation. A trigger checking for new data costs a credit, and nearly every subsequent action, like sending an email or writing to a spreadsheet, costs another. Some AI-related modules consume more than one credit per run.

Is Make.com cheaper than Zapier?

Per unit of work, generally yes for multi-step, moderate-to-high-frequency workflows: Make's Core plan starts at $12/month for 10,000 credits, while Zapier's paid plans start at $19.99/month for 750 tasks. The comparison flips for very simple, low-volume automations, where Zapier's per-task pricing can be easier to predict.

How does Make.com pricing compare to n8n?

They bill on different units entirely. Make charges per module run (credit), so a 20-step scenario costs more than a 2-step one at the same run frequency. n8n charges per full workflow execution regardless of step count, so a complex n8n workflow can cost the same as a simple one. n8n tends to win on cost for step-heavy, lower-frequency workflows; Make tends to win for simpler, high-frequency ones.

What causes Make.com bills to run higher than expected?

Two things most often: polling triggers that check on a fixed schedule and consume credits on every check regardless of new data, and AI or code-execution modules that consume more than one credit per run. Buyers who estimate cost from a simple pilot scenario, without accounting for scheduled polling or planned AI steps, tend to underestimate their real monthly usage.

Is Make.com's $12 price monthly or annual?

Annual. Make's pricing page defaults to the annual toggle, where Core at 10,000 credits is $12/month. Billed month to month, the same plan is $16. The same gap applies at every step of the volume ladder and on Pro and Teams.

What happens when you run out of credits in Make?

Your scenarios stop running until credits are added. Make notifies you at 75% and again at 90% of your purchased credits. Incoming webhooks queue up to your available webhook queue storage and process once scenarios resume, and polling triggers pick up records since their last successful run. Uninterrupted execution past the limit, called overage protection, is an Enterprise feature.

How much do extra credits cost in Make?

Extra credits cost 25% more than the credits included in your plan, whether you buy them manually in bundles of 1,000 or 10,000 or enable auto-purchasing of 10,000-credit blocks. Before November 6, 2025, manual purchases carried no markup and automatic ones carried 30%. Make's own advice is that regular overage buyers should move up a volume tier, where the per-credit rate is lower.

Do any Make modules run for free?

Yes. The Router module consumes no credits, and neither do the error-handler modules Rollback, Break, Resume, Commit and Ignore. Everything else, including triggers that check for data and find none, consumes at least one credit per run.

Does Make charge per user?

No. Make prices by credit volume, not by seat, and its plan comparison table marks unlimited users on all five plans including Free. That is a real difference from Zapier, where the Free and Professional plans are single-seat and multi-user collaboration starts on the Team plan at $69/month for 25 users.

Can I export my Make scenarios and move them elsewhere?

You can export any scenario as a `.json` blueprint from the three-dot menu in the scenario builder, and import it into another scenario or Make account. The blueprint carries modules, settings and mapped values, but not your app connections, which must be reauthorized after import, and the file must be under 2 MB. There is no supported path to convert a Make blueprint into a Zapier or n8n workflow.

What did Make change when it replaced operations with credits?

On August 27, 2025, Make renamed its billing unit from operations to credits and converted existing balances 1:1, leaving plan prices unchanged. One module run still equals one credit for standard apps. The change exists so that AI features and the Code app can meter dynamically, for example at 2 credits per second of code execution, rather than at a flat rate per run.

Is there a published-pricing alternative for connecting AI tools instead of automating workflows?

If the underlying need is giving an AI assistant access to company systems rather than automating workflows between those systems, Coworker MCP publishes its own pricing and connects AI tools to 50+ apps through a single connection, which is a different problem than the one Make, Zapier, and n8n are priced to solve.

Related reading

• Workato Pricing in 2026: The Real Numbers Behind the Quote
• The Best n8n Alternatives
• Best Enterprise AI Workflow Automation Tools