Choose the situation closest to your work. Each note connects a documented limit to a budget decision and a check you can make before buying. These are documentation-based assessments, not hands-on test results.
Zapier
Your Zaps keep running past the allowance
Budget consequence: the subscription fee is not a spending cap. On paid plans, enabled pay-per-task billing adds charges until total task usage reaches three times the plan allowance.
What to verify
Check the extra-task rate and setting in your account; the rate depends on the plan and billing cycle. Pay-per-task is enabled by default for accounts opened after January 2024. Decide whether you prefer extra charges or workflows stopping at the allowance.
Budget consequence: Make’s extra credits cost 25% more than credits included in your plan. Enabled auto-purchasing buys blocks of 10,000; exhausted data transfer can trigger a purchase even when credits remain.
What to verify
Check both credit and data-transfer allowances, then compare occasional extras with a larger allowance for recurring demand. Check expiry too: extras last until the monthly reset on monthly plans and annual Core, or the billing-year end on other annual plans.
Budget consequence: enough monthly executions does not establish enough capacity at peak time. n8n Cloud limits concurrent production executions; in regular mode, additional runs wait in a first-in, first-out queue.
What to verify
Record your busiest minute and acceptable delay, not just monthly volume. Run a controlled burst against the proposed plan and inspect how long work waits. A higher execution allowance is useful only if its concurrency limit also meets the job.
Budget consequence: replaying an entire Zap creates a new run. Successful billable steps use tasks again, including steps that succeeded before the failure.
What to verify
Compare retrying errored steps with replaying the entire run. Before repeating a create or send action, inspect what reached the destination. Include recovery usage in the budget and test how duplicate records or messages are prevented.
Budget consequence: a custom AI-provider connection can mean Make credits plus a separate provider token bill. Make’s own AI Provider instead consumes credits based on operations and tokens; an AI step is not necessarily one credit.
What to verify
Identify the connection type and model for each AI module. Price a representative input and output, then record both Make usage and any provider charges. Do not multiply the ordinary one-credit rate by every AI step.
Budget consequence: free self-hosting does not include every team feature. n8n Community excludes Projects, workflow and credential sharing, and SSO. A shared operational setup may need a paid edition as well as hosting.
What to verify
Have a second authorized operator inspect and recover a sample workflow using their own account. Check the required sharing and sign-in features against the proposed edition. Free Community registration does not unlock those collaboration features.
These are historical observations from , not a live price feed. The current price, billing period, allowance and required features need to be checked together before you buy.
Zapier — Tasks
Free includes 100 tasks per month. Paid plans scale with the task allowance.
Professional from USD 29.99/month billed monthly, or USD 19.99/month billed annually. Team from USD 103.50/month monthly, or USD 69/month annually.
The capture records USD and separate billing options. Trigger billing remains unresolved: the rates page describes standard triggers as chargeable, while the task-usage help page excludes triggers. Both were checked ; this is not evidence of a confirmed billing change.
Free includes 1,000 credits per month. The paid-plan display uses 10,000 credits.
At 10,000 credits: Core $16, Pro $28, Teams $51/month billed monthly. Annual equivalents: $12, $21, $38/month.
The capture separates billing options but does not establish the dollar currency code. It is not a verified USD comparison. Modules with weighted usage can consume more than one credit.
Starter includes 2,500 executions; Pro includes 10,000. Both advertise unlimited steps.
Starter $24/month and Pro $60/month billed monthly; annual equivalents $20 and $50/month.
The historical dollar capture does not establish its currency code. n8n’s billing guidance, checked , says US billing addresses pay in USD, UK addresses in GBP, Brazil in BRL, and most others in EUR. Pre-checkout display depends on browsing location; confirm the final currency and amount for your billing country.
No cheapest-tool verdict from these figures. Currency and task-accounting gaps prevent a like-for-like price ranking. A newer capture alone would not prove a price change: the plan, allowance, currency and billing period must match. Use the calculator to inspect modeled usage, or compare documented requirements and your own quotes.
Use the same synthetic records and required outcome across your shortlist. A lower counter is not a saving if one tool skips work the other completes.
Define done. Name the destination record, required fields and acceptable delay.
Include a failure. Repeat a record, omit a field and interrupt a connection. Check the destination, not only the success indicator.
Keep the receipt. Record plan, date, completed items, billed units, retries, external charges and operator time. Separate measured usage from any monthly projection.
These are proposed trial checks, not results we have measured.