Make.com Pricing in 2026: The Credit Ladder, AI Model Costs, and the Polling Checks You Pay For

How much does Make.com cost? The short answer

As of October 6, 2026, Make is free for 1,000 credits a month. Core costs $12 a month billed yearly ($16 billed monthly) for 10,000 credits, Pro $21 ($28 monthly), and Teams $38 ($51 monthly) at the same 10,000 credits; Enterprise is quoted. Most module actions use 1 credit, AI steps can use more, and a scenario that polls for new data spends a credit on every check, even when it finds nothing. That last rule can matter more to your bill than your actual volume.

Make renamed its billing unit from operations to credits; its help center says existing plans and credit prices did not change. Every number here is from Make's pricing page (including the price ladder the page loads for its credit selector) and two Make Help Center articles, all read October 6, 2026.

The plans at 10,000 credits

The credits are the same on every paid plan: Pro and Teams charge more for features, not for a better credit rate. At 10,000 credits, Teams billed monthly is more than three times Core for the same amount of automation.

The full credit ladder

The plan cards show 10,000 credits. The selector goes much higher. Core stops at 300,000 credits a month; Pro and Teams run to 8 million.

Yearly prices are per month, paid up front. The page advertises "Save 15% or more" for yearly billing; at the 10,000-credit Core tier the gap is 25%. Yearly billing can also change when credits expire: Make's FAQ says credits on annual Pro and Teams subscriptions last to the end of the year instead of resetting each month. On monthly billing, unused credits expire at the end of the month, and Core is not named in that rule, so confirm before you count on rollover.

What uses a credit

Make's FAQ says each module action uses a credit: reading, searching, creating, updating or deleting data in an app, transforming data with the built-in tools, and aggregating or iterating rows. A run can use "anywhere from two credits to thousands."

• Router and error handler modules are free. Rollback, Break, Resume, Commit and Ignore use no credits.
• Each bundle multiplies the modules after it. A trigger that returns 10 new form responses into a 3-module scenario uses 1 + 30 = 31 operations, in Make's own example.
• A trigger uses one operation per check. Make's operations article gives the example directly: a Watch New Rows trigger "checks for new rows in a Google spreadsheet = 1 operation (one per check)." The check costs the same whether it finds ten rows or none.
• The Code app uses 2 credits per second of execution time.
• AI depends on how you connect. With your own OpenAI or Anthropic key, an AI module is 1 credit per operation and you pay the provider for tokens. With Make's AI Provider, credits are based on tokens and operations, at published rates per model.

The polling check you pay for when nothing happens

This is the rule that moves Make bills the most, because it has nothing to do with how much work gets done. A scenario that watches an app on a schedule pays one credit every time it looks:

One scenario polling every 15 minutes uses up the Free plan's 1,000 credits in about ten and a half days with no data at all. An instant webhook trigger avoids this entirely: it runs only when the other app sends something.

A worked example: 300 leads a month

A home-services company gets 300 web leads and 150 missed calls a month. Its lead scenario takes the form entry, uses an AI module (with the company's own model key) to score the lead, creates the CRM contact, texts the lead, adds a follow-up task and posts to the team chat: 6 credits per lead. A missed-call text-back uses 2 credits per call. A daily summary uses about 3 credits a day. A chaser scenario searches the CRM for leads nobody has called back and sends about 60 reminders a month.

The leads, the calls and the messages are identical in every row. The bill runs from $12 to $129 a month on yearly billing because of two trigger settings. If a setup feels like it needs minute-by-minute polling, that is usually a sign the app offers a webhook you have not used yet.

What the AI model does to the same lead

Swap the company's own model key for Make's AI Provider and the AI step is billed in tokens. Make publishes how many tokens one credit buys for each model. For a lead prompt of about 1,500 input tokens and a 300-token answer, our arithmetic on those published rates gives:

At 300 leads, the Opus choice alone is about 5,000 credits a month, more than every other step in the scenario combined. Sorting and summarizing leads rarely needs the largest model. Make notes that its models and conversion rates "are subject to change," so check the table in its help center before you budget.

What happens when you run out

On Core, Pro and Teams your scenarios stop until credits are added. Make sends a notification at 75% and again at 90% of your credits. Incoming webhooks queue up to your webhook queue storage and process when you resume, and polling scenarios pick up records from their last successful run, so a pause delays work rather than losing it, as long as the queue holds. To keep running you can upgrade, buy extra credits in bundles of 1,000 or 10,000 at the price set in your subscription, or turn on auto-purchase of 10,000 credits before you run out. Uninterrupted running past the limit, called overage protection, is an Enterprise feature.

Make vs n8n vs Zapier on the same lead

The three tools meter the same lead workflow three different ways. Make counts module actions, about 6 credits for this lead. n8n counts the whole run as one execution, and our n8n pricing breakdown shows why its schedule triggers, not its leads, set the plan. Zapier counts successful steps, with an AI step at 1, 3 or 5 tasks by model tier, which is the subject of our guide to how much Zapier costs. For a wider field of tools priced around AI steps, see the best AI workflow automation tools.

Is Make worth it for a small business?

For webhook-driven workflows at small volume, Make is one of the cheapest ways to run real automation: the 300-lead setup above costs $12 a month on Core billed yearly. It gets expensive in two situations: scenarios that poll on short intervals, and AI steps run through the largest models on Make's AI Provider. Both are design choices, which is why the same business can pay ten times more for the same result.

Who can build a Make automation for my business?

You can build it yourself, hire a freelancer, or hire a studio that designs the whole system. Our guide to AI automation pricing lays out what each route typically costs.

SuperDupr builds these systems for service businesses from Austin. Our published price for connected AI workflow automation is $10,000 to $25,000, with $200 to $2,000 a month ongoing depending on volume, and most builds go live in 2 to 4 weeks. The build is owned by your business, and the platform underneath, Make, n8n or another, is chosen by what it costs at your volume, including how the triggers are set. You can book a free 30-minute strategy session from that page.

Sources

• Make pricing page, make.com/en/pricing, read October 6, 2026: plan cards, comparison table, FAQ, and the price ladder loaded by the page's credit selector.
• Make Help Center, "Credits," help.make.com/credits, read October 6, 2026: how credits are used, AI connection types, and the tokens-per-credit table for Make's AI Provider.
• Make Help Center, "Operations," help.make.com/operations, read October 6, 2026: operations per trigger check and per bundle.
• n8n pricing and execution documentation, and Zapier's task usage rates page, read October 6, 2026, for the comparison section.