Make vs n8n: what 4,000 orders a month cost

A small online shop syncs every new order to its accounting tool and a spreadsheet. The workflow has a trigger and 6 action steps (fetch the order, create the invoice, update the sheet, tag the customer, send the confirmation, notify the team). It runs about 4,000 times a month. The trigger checks for new orders every 5 minutes. This guide counts this one workflow only; if your account runs several, count each one the same way and add the totals. You can change any of these numbers in the calculator.

Plan details for each platform are on its product page: Make and n8n Cloud. For how each platform counts usage, see how automation pricing works.

The trigger alone is just over a quarter of Make's count here (8,640 of 32,640). n8n's number does not depend on the 6 steps at all.

On the pages as they appeared to us, Make is priced in US dollars and n8n in euros (n8n's price feed also carries dollar, sterling and real prices for visitors from some countries, so what you see may differ); they are not converted, and a price on one billing basis is never compared with a price on the other. Read each row against itself, not as “$ is cheaper than €”. Make sells fixed credit sizes (its slider offers fixed credit sizes (we saw 40,000, 80,000, 150,000 and 300,000 on Core; larger sizes are offered on Pro and Teams only)), so a count just above a size moves to the next one.

The 15-minute row counts 2,880 checks (every 15 minutes × 30 days) plus 4,000 runs × 6 actions = 24,000 credits, which is 26,880 credits in total. The loop row is the one that matters. In Make each module runs once per item, so ten items per run multiply the action credits by ten. In n8n a loop inside a workflow is not counted: still one execution. One caveat: Make's pricing FAQ also lists iterating and aggregating rows among the actions that use credits; this calculator counts only the action steps you enter, so add them yourself if your scenario loops through an iterator or aggregator. Routers and error-handler modules use no credits on Make, so they do not add to these counts (Make's pricing FAQ). At the extreme, a nightly job that pushes 200 products through 4 steps, only 30 runs a month, is 24,030 credits on Make (Core, $29/mo) against 30 executions on n8n (Cloud Starter, €20/mo).

How we counted Make credits, and where the count stops. For the main scenario the formula is: credits per month = trigger checks (check interval × 30 days, 8,640 at 5 minutes) + runs × action modules per run (4,000 × 6 = 24,000), which is 32,640. It assumes one workflow, a 30-day month, one credit for each trigger check and for each action module run, no AI modules, no iterators or aggregators (the loop row above shows what several items per run add) and no extra credits bought. Make's help pages say that a trigger uses one operation per check and each module run uses one credit; routers and error-handler modules use none. We could not confirm from Make's documentation how every individual trigger type or app module is billed, so a module that bills differently would change the total. The plan shown for each row is therefore an estimate for the scenario described here, not a price for any similar workflow.

• Plan limits. Make supports scheduling intervals as short as one minute, but the minimum interval available depends on the plan. We could not verify the current active-scenario limits for Core from official documentation. n8n Cloud Starter allows 5 concurrent executions, and Pro is shown as “up to 50” on its card and “20 / 50” in n8n's comparison table, which also lists a maximum execution duration of 5 minutes on Starter and 40 minutes on Pro. A 200-item loop is cheap on n8n, but check that one run fits the limits of the plan you would buy.
• Polling is mostly a Make cost. On Make every check of a polling trigger is an operation. On n8n a polling trigger counts only when it finds new data, while a schedule trigger counts every firing (n8n docs). With an instant trigger (webhook) there are no idle checks; the calculator then counts one credit per run, which is its assumption rather than a rule stated on the pages read.
• Overage. Make lets you buy extra credits in bundles of 1,000 or 10,000 at the fixed price of your subscription, or upgrade (Make's pricing FAQ); n8n lists overage fees as capped for Starter and Pro but publishes no amount. We modelled neither: the tables show the next size up instead.
• Self-hosting n8n removes the licence fee, not the cost: see the guide on self-hosting economics.

• A workflow with no loop and no frequent polling (say 300 runs a month, 2 steps) fits Make's free plan (1,000 credits, 15-minute minimum interval): 900 credits. Neither paid plan above is needed.
• More than 50,000 executions a month on n8n Cloud: Cloud Pro's largest published size is 50,000. The next published n8n plans are the Business licence sizes (from 40,000 executions), which are self-hosted only.
• Planning around spikes: Make's price for extra credits depends on your subscription and n8n publishes no Cloud overage price, so spikes cannot be priced from the pages read here.

Both vendors change plans and credit or execution sizes. Count your own workflow first (the first step above), then check the size and billing basis on each vendor's own pricing page: Make is priced in credits per month, n8n Cloud in workflow executions per month. The two are listed alphabetically; the order does not reflect any commercial relationship. A link marked “Partner link” may earn us a commission; a link without that label goes to the vendor's official site. This does not change the calculations or our conclusion.