Using AI in Make is not, by itself, a reason to move from Core to Pro. Make’s AI agents and toolkit appear on every plan, and since November 2025 Core can connect your own AI provider. If you only need more credits, raise the Core slider or buy extra credits. Pay for Pro when you need custom variables, full-text log search, larger files, higher API limits, priority execution or one annual credit pool.
Pricing and plan comparison read in a browser on October 31, 2026, in US dollars. The seasonal example below is invented. This article contains no affiliate links.
Separate capacity from features
Make sells credits on a slider and plans as feature sets. Most “should I upgrade” questions mix the two. The table puts them side by side at the same 10,000 credits a month, so that the difference you see is only the plan.
| Decision point | Core | Pro |
|---|---|---|
| Price billed annually | $9 a month, $0.90 per 1,000 credits | $16 a month, $1.60 per 1,000 credits |
| Price billed monthly | $10.59 | $18.82 |
| Largest monthly credit allowance | 300,000 | Up to 8 million |
| Unused credits on annual billing | Allowance is monthly; credits expire each month | Annual credits that “only expire after 12 months instead of each month” |
| Priority scenario execution | Not included | Listed as “High” |
| Custom variables | Not included | Included |
| Full-text execution log search | Not included | Included |
| Maximum file size | 100MB | 250MB |
| Make API rate limit | 60 calls a minute | 120 calls a minute |
| The same on both | Unlimited active scenarios, 1-minute scheduling, 40-minute maximum run, 30 days of execution logs, code at 2 credits per second, and the Make AI rows: AI agents, AI toolkit, MCP server | |
Sources: Make pricing, both billing toggles and the comparison table, and Make’s help-center notice on plan and pricing adjustments dated November 6, 2025 for the credit limits. Prices move with the slider: at about 17,500 credits the page showed Core at $14 and Pro at $26 billed annually, so the price per 1,000 credits falls as you buy more on either plan. Pro costs roughly 1.8 times Core at both settings we read.
If you are still deciding whether to leave the Free plan at all, our Make Free vs Core operations math covers that step. This article starts where that one stops.
Where the AI cost actually goes
For an AI workflow, the plan matters less than how you call the model. Make’s help center describes three routes. A third-party AI app module, such as OpenAI, Claude or Gemini, uses one credit per operation like any other module. Make’s own AI Provider charges credits based on tokens and operations, so a long prompt costs more credits. And a custom AI provider connection means Make charges only the operation, while you pay the model provider for tokens directly.
That last route used to be Pro-only. Make’s November 6, 2025 notice says custom AI provider connections are now available on all paid plans, including Core, for features such as Make AI Agents and Make AI Toolkit. If a guide written before then told you to buy Pro for your own OpenAI or Anthropic key, that reason no longer applies.
So the first question for an AI enquiry workflow is not “Core or Pro” but “which AI route, and how many credits a run”. Our guide to Make AI steps for routing client enquiries walks through building that kind of scenario; count the modules in yours and multiply by your monthly volume before looking at plans.
When Pro’s features earn their price
- Custom variables. If several scenarios share the same values, such as a prompt version, a routing threshold or a client’s folder ID, Pro lets you edit them in one place. On Core you edit each scenario.
- Full-text execution log search. When a client asks why one enquiry was classified the wrong way, searching run history for their email address is faster than opening runs one by one. Both plans keep 30 days of logs; only Pro searches them.
- Files over 100MB. Contract bundles, recordings or large spreadsheets sent through an AI extraction step can exceed Core’s 100MB limit. Pro raises it to 250MB.
- The Make API at volume. If you manage scenarios through Make’s own API, Pro doubles the rate limit from 60 to 120 calls a minute.
- Priority execution. Make describes it as running scenarios “during peak usage times to minimize queuing time”. It publishes no waiting-time figures for either plan, so treat it as a tiebreaker rather than a reason.
The seasonal-volume test
Pro’s annual credit pool is the one feature that is really about capacity. On annual Core, each month’s credits expire at the end of that month. On annual Pro, a year’s credits can be used in any month. That matters if your enquiries arrive in bursts, such as a launch, a tax deadline or a holiday season.
Make’s help center says extra credits cost 25% more than the credits in your plan, bought in multiples of 1,000; the pricing page FAQ describes them as sold at the fixed price defined in your subscription. The example below uses the help-center rate. It is an invented business: ten quiet months at 6,000 credits and two busy months at 22,000, a yearly total of 104,000.
| Line | Core plus extra credits | Pro with annual pool |
|---|---|---|
| Subscription for 12 months | $108 (12 × $9) | $192 (12 × $16) |
| Credits short in the two busy months | 24,000 (12,000 each) | None: 104,000 fits in the 120,000 pool |
| Extra credits at $0.90 per 1,000 plus 25% | About $27 | $0 |
| Approximate year | About $135 | $192 |
In this example Core is cheaper even after buying extra credits. The price gap at this setting is $84 a year, which buys roughly 75,000 extra credits at the rate above. The pool only wins when your busy months are short by more than that in a year, which takes strongly concentrated demand. Run the same sum with your own months before paying for Pro on the strength of the pool. Remember that extra credits bought on a monthly subscription expire after a month, and unused plan credits expire at the end of the term on both plans.
What happens when credits run out
On either plan, Make’s FAQ says “Your scenarios won’t continue to run, until credits are added again.” Incoming webhooks are queued up to your webhook queue storage, which scales with the credits you license, and polling scenarios look for records since their last successful run once they resume. For a client enquiry form that posts to a webhook, that queue is your safety net; check it is large enough for a busy day. Core, Pro and Teams can also turn on automatic purchase of 10,000 extra credits before the balance runs out.
Upgrading, downgrading and the trap in between
- Measure first. Look at last month’s credit use per scenario and list which Pro features you would actually use. If the list is empty, you need a bigger Core slider, not Pro.
- Try monthly before annual. Monthly billing costs more per credit but lets you reverse a wrong choice after one cycle.
- Know the downgrade rule. Make’s FAQ says that if you downgrade, “any active scenarios using features of higher plans will be stopped.” Scenarios rebuilt around custom variables will stop on Core, so note which ones use Pro features.
- Keep a human check on AI output. Neither plan reviews what the model wrote. Route low-confidence classifications or empty fields to a review step before anything reaches a client.
Which plan fits
| Your situation | Plan |
|---|---|
| Steady volume, AI through a third-party module or your own key | Core, with the slider set to your usage |
| Occasional busy months | Core plus extra credits, after running the seasonal test |
| Most of the year’s demand in one or two months | Pro, for the annual pool |
| Many scenarios sharing prompts or settings | Pro, for custom variables |
| Frequent client questions about individual runs | Pro, for log search |
| Files over 100MB in an AI extraction step | Pro |
| A team that needs shared roles and templates | Teams, which neither Core nor Pro covers |
Frequently asked questions
Do I need Pro to use my own OpenAI or Anthropic key?
No. Make’s November 6, 2025 notice says custom AI provider connections are available on all paid plans, including Core.
Does Pro keep execution logs longer?
Not according to the comparison table we read: Core and Pro both show 30 days. Pro adds full-text search across those logs.
Is Pro faster?
Pro has priority execution, which Make says minimizes queuing at peak times. Make publishes no figures.
What happens to my extra credits if I downgrade?
Make’s FAQ says unused credits transfer to the new subscription and remain valid for the next month or year, depending on billing.
Sources and method
Prices, the comparison table and the FAQ were read on Make’s pricing page in a browser on October 31, 2026, at the 10,000-credit setting on both billing toggles and at one higher slider setting. Plan credit limits, the extra-credit surcharge and custom AI provider availability come from Make’s November 6, 2025 adjustments notice, and AI credit rules from its credits documentation. The seasonal table is our arithmetic on invented volumes; it is not a quote. Before you rely on it, check queue times with and without priority execution, the exact charge for extra credits after tax and the credit cost of your own scenarios.
Read next
