Automation capacity guide

n8n execution limit: count and reduce runs

If your n8n Cloud Pro plan says the execution limit has been reached, first find out which workflows are starting most often. An execution is one complete workflow run, not one individual node operation. A schedule that starts too often, a published workflow still being tested, or unnecessary polling can use the allowance faster than expected.

How the n8n execution limit works

n8n Cloud Starter includes 2,500 executions per month, and n8n Pro starts at 10,000 executions per month; n8n's docs also list a 50,000-execution Pro tier. n8n Business is self-hosted only and includes 40,000 executions. One execution is one full workflow run regardless of how many steps it contains. That makes start frequency a useful first diagnostic: adding steps to one run does not by itself create a separate execution, but triggering the workflow more often does.

Checked October 2026, confirm on the vendor's pricing page.

Scheduled triggers can add up quickly. For example, 10 workflows each running every 5 minutes would create 2,880 executions in a day if every scheduled run occurs. n8n estimates that a 5-minute schedule is about 8,600-8,900 executions per month. Check actual usage in your execution history.

n8n community members report that workflows stop when the monthly limit is reached and that extra executions can't be bought on Starter or Pro; upgrading is the usual fix. Confirm with n8n. n8n says Business and Enterprise workflows keep running and overage charges may apply. n8n Cloud has no permanent free plan, but offers a free trial (no card required), and the Community Edition is free to self-host. The self-hosted Community Edition has no execution cap, but you are responsible for operating the server.

How to lower usage before changing plans

Start with measurement, not guesswork. Sort execution history by workflow and compare the number of starts with the work each run accomplishes. Note the trigger, the schedule, and whether a run is useful. Then make one measured adjustment at a time so you can see whether it changes volume without delaying work that matters.

1. Replace polling with webhooks when suitable. A webhook can start work when an event happens rather than checking repeatedly for a change. Use polling only when a source cannot notify you or when its delay is acceptable.
2. Relax schedules. If a workflow checks more frequently than its business task requires, lengthen the interval. Coordinate with the people relying on the result before changing expected response times.
3. Unpublish test workflows. Keep experiments from running on a production schedule. Test deliberately, then disable or remove triggers that are no longer needed.
4. Combine related checks thoughtfully. A single scheduled run can inspect several related conditions, but avoid building a large, fragile workflow just to reduce a counter. Maintain clear failure handling and ownership.
5. Estimate upcoming load. Multiply expected starts per day by operating days, and include retries or multiple active copies only after confirming how your implementation behaves. Compare that estimate with observed history.

When another plan is worth comparing

If your well-tuned workloads still exceed the allowance, compare capacity, node limits, workflow limits, hosting responsibility, and the cost of changing tools. iii3xnz Starter is published at $15 per month with 20,000 executions monthly, up to 25 workflows, and up to 60 nodes per workflow. Its free plan has 250 executions per day; its Pro plan has 150,000 per month for $99. All plans expose the same node types, so the published distinction is the plan capacity and other account features, not a plan-specific node gate.

These product limits are listed in the frontend plan contract and enforced in backend plan limits. They are not a promise that every workflow will be a fit: check the available nodes, credentials, and operational requirements for your actual scenario before migrating. If you need n8n’s particular self-hosting model, its Community Edition may be a better match despite the server work.

Where n8n is better

n8n may be the better choice if you value its workflow environment and can host Community Edition yourself, or if its specific nodes and deployment model already fit your team. Self-hosting avoids a Cloud execution cap but shifts server operation to you. n8n Cloud keeps the hosting managed, while its usage allowances are plan-based. Decide based on the cost of your team’s time, the trigger pattern, and the services your workflows need—not only on a single monthly limit.

Frequently asked questions

How does n8n count an execution?

On n8n Cloud, one execution is one complete workflow run, whether it contains a few steps or many. Repeated scheduled starts can therefore use the allowance quickly.

What can I do when my n8n Pro limit is reached?

Review run history, reduce unnecessary polling, use webhooks where they fit, and avoid publishing test workflows. n8n community members report that workflows stop when the monthly limit is reached and that extra executions can't be bought on Starter or Pro; upgrading is the usual fix. Confirm with n8n.

Does adding nodes increase the n8n execution count?

No. The Cloud count is based on complete workflow runs, not the number of steps inside each run. More frequent starts, rather than more nodes alone, raise the run total.

What does iii3xnz Starter include?

The published Starter plan is $15 per month and includes 20,000 executions per month, 25 workflows, and up to 60 nodes per workflow. Check the product pricing before choosing a plan.