n8n Pricing, Explained Simply

If you have shopped for workflow-automation tools before, you have probably been burned by confusing pricing. Most platforms meter you on operations or tasks every single action inside an automation ticks a counter, so a workflow that touches ten apps quietly costs ten times as much as one that touches one. n8n takes a deliberately different approach, and understanding that difference is the key to understanding the whole price list.

On n8n, the unit you pay for is the execution: one complete run of a workflow from its trigger to its final node. It does not matter whether that workflow has two steps or two hundred, whether it loops over a single record or ten thousand it counts as exactly one execution. That single design decision makes n8n dramatically cheaper for the data-heavy, multi-step, AI-driven automations that technical teams actually build, and it makes your monthly bill far easier to predict. The sections below walk through each plan, what an execution really means, how AI Workflow Builder credits fit in, how the self-hosted Community Edition stays free forever, and how to right-size your plan before you ever enter a credit card.

An execution is a single run of your entire workflow, triggered to completion. It is the single most important concept in n8n pricing, because it is the only thing you are billed on. Competing platforms typically bill on operations or tasks, where each node, action, or API call inside an automation increments a meter. On those tools, a workflow that fetches a list of 500 customers and updates each one in a CRM might burn 500+ operations in a single run. On n8n, that same run is one execution.

This is why n8n usage is so much easier to forecast. You are not trying to predict how much data will flow through a workflow or how many branches it will take you are simply counting how many times each workflow fires in a month. The complexity lives inside the execution, where it is effectively free; the meter only ticks once per run.

Why this matters for your bill

Consider a realistic data-enrichment pipeline: a webhook receives a new signup, the workflow calls an enrichment API, runs an AI model to classify the lead, writes to a database, posts to Slack, and sends a welcome email. That is six or more billable operations on a per-task tool, every single time it runs. On n8n it is one execution. Multiply across thousands of signups a month and the gap between "per task" and "per execution" becomes the difference between a four-figure and a two-figure invoice.

How many executions does each plan include?

The published monthly allowances are straightforward: Starter includes 2,500 executions, Pro includes 10,000, and Business includes 40,000. The Enterprise plan is sized to a custom number negotiated with sales. Importantly, only production executions count toward your quota in the way you would expect, and active workflows are unlimited on every paid tier there is no cap on how many automations you can keep switched on, only on how many times they collectively run.

What happens if you go over?

If you exceed your quota on Business or Enterprise, your workflows keep running without interruption n8n does not shut you off mid-automation. Instead, overage charges may apply if you do not move up to the next usage tier. When you approach your limit, the sales team reaches out to discuss your usage and the right tier. If there is no response or upgrade, overages are invoiced after the fact; for the Business plan the published overage rate is €4,000 for each additional bucket of 300,000 executions. Annual Business subscribers also get a heads-up from the n8n team when they hit roughly 80% of their annual quota.

n8n bundles a feature called the AI Workflow Builder, which generates and edits workflows for you from natural-language prompts. This is metered separately from executions, using its own pool of credits, and the monthly allowance grows with your plan:

• Starter: 50 AI Workflow Builder credits per month
• Pro: 150 AI Workflow Builder credits per month
• Enterprise: 1,000 AI Workflow Builder credits per month (available on n8n Cloud)

A crucial distinction: these AI Builder credits are not the same as the cost of running AI inside your workflows. When you add an AI Agent node or call an LLM such as OpenAI, Anthropic Claude, or Google Gemini as part of an automation, that usage is billed directly by your model provider through your own API key it does not consume AI Builder credits. The credits only cover the assistant that helps you construct workflows, not the AI steps that run inside them. During a free trial you receive a smaller allotment (20 AI Workflow Builder credits) so you can try the feature before committing.

You can try n8n Cloud before you pay. When you start a free trial on the Starter or Pro plan, you actually receive access to all Pro-tier features during the trial window including Global Variables, Insights, and Admin Roles so you can evaluate the full toolkit rather than a stripped-down version. The trial operates within Starter-level resource limits: 5 concurrent executions, 320 MiB of RAM, 10 millicore burstable CPU, and 2,500 executions, plus 20 AI Workflow Builder credits.

For the Business plan, the free trial gives you full access to all Business features for 14 days. The key practical difference is the payment method: no credit card is required for Starter and Pro trials, whereas a credit card is required to begin the Business plan trial. This lets individuals and small teams explore the product risk-free, while still giving larger organisations a genuine hands-on evaluation of the collaboration and governance features they are buying.

This is one of the most common questions about n8n pricing, and the answer reveals the philosophy behind the whole model. The free Community Edition is genuinely free and genuinely unlimited you can run it forever without paying n8n a cent. What the paid Business and Enterprise plans add, even on self-hosted instances, are the capabilities that companies driving measurable outcomes with automation tend to need: SSO/SAML/LDAP, multiple environments, Git-based version control, audit logging, external secret stores, log streaming, and dedicated support backed by an SLA.

n8n is explicit that its pricing is tied to the value automation delivers to your organisation, not to what it costs n8n to run the software. The reasoning is that teams operating at the Business or Enterprise level are usually saving substantial time, streamlining operations, and sometimes unlocking new revenue so the licence pays for itself many times over. To unlock those paid features on a self-hosted instance, you apply a licence key that pings n8n's licence server daily to stay active. That ping reports usage data such as production execution counts, which is how n8n meters self-hosted Business and Enterprise usage. You can apply a single licence key across an unlimited number of instances, and the combined usage from all of them counts toward your quota.

On data residency: for n8n's hosted plans, your data is stored within the EU, on servers located in Frankfurt, Germany. For self-hosted deployments, your data lives wherever you choose to run n8n which is precisely why regulated industries with strict residency requirements often prefer the self-hosted route.