Make.com Pricing Calculator

Operations are now credits

If you learned Make's pricing before November 2025, the vocabulary has changed. Make renamed its billing unit from operations to credits, and the two words now mean different things: credits track what you spend, operations describe what happened on the platform.

For most work the arithmetic survives the rename. One module action equals one credit for non-AI apps, and also for third-party AI apps like OpenAI or Anthropic where you supply your own API key, in which case you pay Make for the credit and your provider for the tokens.

What did not survive is the assumption that every module costs the same. Make now splits credit usage into two kinds:

• Fixed usage. A set rate per run, regardless of input size. Uploading a file through Google Drive costs one credit whether the file is 5 KB or 500 MB. A handful of modules carry a fixed rate above the default and are tagged in the interface with the number, such as 2 or 10 credits per operation.
• Dynamic usage. The rate varies with actual consumption, driven by tokens, file size, page count or processing time. These modules carry tags in the builder telling you which factor applies.

Plans and allowances

Figures below were read from Make's pricing page on 13 August 2026, at the base 10,000-credit position on the slider. The page carries a monthly and annual billing toggle advertising a saving of 15% or more, so confirm which view you are looking at before you commit a budget.

The credit slider steps through 10k, 20k, 40k, 80k, 150k, 300k, 500k, 750k, 1M and upwards, and the price moves with it. Data transfer scales alongside at roughly 5 GB per 10,000 credits a month, against 512 MB on Free. Maximum file size runs 5 MB on Free, 100 MB on Core, 250 MB on Pro and 500 MB on Teams.

Worked example: an order-processing scenario

A Shopify store runs one scenario on every new order: look up the customer in the CRM, create or update the contact, append a row to a Google Sheet, send a confirmation email, post to Slack, create a fulfilment task, adjust inventory, write an audit log. That is eight module actions, so eight credits per order.

At 1,500 orders a month that is 12,000 credits, which overshoots the 10,000 Core allowance by 2,000.

Option one, extra credits. Since 6 November 2025, extra credits cost 25% more than the credits included in your plan, whether you buy them manually in multiples of 1,000 or let auto-purchase pick up 10,000 at a time. Core at $12 for 10,000 credits works out at $0.0012 a credit, so the top-up rate is $0.0015. Two thousand extra credits is about $3, taking the month to roughly $15.

Option two, move up the slider. Make states plainly that higher tiers carry a lower per-credit cost. At a steady 12,000 credits a month, the 20,000 position will almost certainly beat $12 plus permanent top-ups. Extra credits are a buffer for a spike, not a plan.

Where it actually blows up. Add an iterator that loops over each line item and runs three modules per item. An order with twenty items no longer costs three credits at that stage, it costs sixty. One innocuous loop can multiply a scenario's cost by an order of magnitude, and nothing on the canvas warns you.

What an AI module really costs

This is where the old operations mental model breaks hardest. If you use Make's own AI provider rather than your own API key, credits are consumed against a published token-to-credit rate. On gpt-5.6-terra that is 452 input tokens per credit and 75 output tokens per credit, halving again once the context passes 272,000 tokens.

Take a 3,500-token prompt returning 250 tokens. That is roughly 7.7 credits of input plus 3.3 of output, so about 11 credits for a single AI step — the same as eleven ordinary modules. At Core's $0.0012 a credit, around $0.013 a run.

Whether that is good value depends on what you are comparing it to. Calling OpenAI directly on the same model costs about $0.010 for that request at list rates, and closer to $0.005 with prompt caching configured properly, but you then manage the key, the retries and the caching yourself. The OpenAI API Cost Calculator works that side of the comparison through in detail.

Common mistakes

Budgeting from the module count instead of the run count. Eight modules is not eight credits a month, it is eight credits every time the scenario fires. Multiply by frequency first, then by loop depth.

Assuming AI modules are one credit. They are dynamic. Check the tag on the module before you build a scenario that runs it a thousand times a day.

Living permanently on extra credits. A 25% premium on every credit above your allowance is fine for a seasonal spike and expensive as a habit.

Testing on Free and costing the build from it. Free caps you at two active scenarios and a fifteen-minute minimum interval, so anything near real-time behaves differently the moment you go live.

Forgetting data transfer and file size. A scenario that moves large attachments can hit the file size ceiling on a plan whose credit allowance is comfortable.

Frequently asked questions

Is it still called operations? Both words are still in use, but they are not interchangeable. Credits are what you are billed in; operations describe activity. For non-AI modules one operation still equals one credit, so old scenario estimates convert directly.

How do I count credits before I build? Build the scenario once, run it against ten realistic records, and read the actual credit consumption from your organisation's usage view. Estimating from the canvas reliably undercounts, because loops and error paths are invisible until they run.

What happens when I run out mid-month? Scenarios stop until you upgrade, buy extra credits, or the cycle resets. Auto-purchase avoids the outage by buying 10,000 credits at the 25% premium, which is worth enabling on anything customer-facing and worth reviewing monthly.

Is Make cheaper than Zapier or n8n? It depends entirely on shape. Make bills per module action, so deep multi-step scenarios are where it wins; Zapier bills per task, which favours short two-step automations. The Zapier Cost Calculator runs the task-based maths, and the No-Code Tool Comparison Matrix lines the platforms up feature by feature. Make.com vs Zapier for Small Business works through where each one actually costs less.

I am new to Make. Where should I start? Make.com Tutorial for Beginners covers building a first scenario end to end, including the error handling most tutorials skip.

Getting the estimate right before you build

Most Make bills that surprise people are not a pricing problem, they are a design problem: a loop that should have been a bulk call, an AI step doing work a filter could do, or a scenario polling every minute when every hour would do.

If you want a scenario costed and tightened before it goes live, I take that work through Fiverr and Upwork.