n8n Pricing (2026): Plans, Executions, and How to Pick the Right Tier

How we checked: prices, execution allowances, and plan features come from n8n's official pricing page, its Start-up plan page, the n8n Docs, and the n8n Help Center, all checked on October 16, 2026 (sources at the end). We have no affiliate relationship with n8n. The pricing page showed prices in euros for us. The Help Center quotes add-on prices in euros, dollars, and pounds, so the currency you're billed in may depend on your region.

The short answer

n8n charges per workflow execution, not per step. One run of a workflow counts once, however many nodes it has and however much data it moves. That makes it very different from per-task or per-operation pricing.
Starter (€24/month, or €20/month billed annually) gives you 2,500 executions a month. That is enough for a handful of hourly or daily workflows. It is not enough for a single workflow on a 15-minute schedule (see the math below).
Pro starts at €60/month (€50/month annually) for 10,000 executions, with a larger 50,000-execution tier at €145/month. Most people running real production automations end up here.
Business (from €800/month, or about €667/month annually) is self-hosted only. You pay for team features like SSO, Git version control, and multiple environments, not for hosting.
Community Edition is free if you host n8n yourself and don't need the paid features.

Plans side by side

Euros, from n8n's pricing page on October 16, 2026. "Annual" is the yearly price. n8n advertises annual billing as a 17% saving, which matches the numbers below.

Plan	Where it runs	Monthly billing	Annual billing	Executions / month	Concurrent executions
Community Edition	Your server	Free	Free	No n8n plan quota	Set by your own setup
Starter	n8n Cloud	€24/mo	€240/yr (€20/mo)	2,500	5
Pro (10k)	n8n Cloud	€60/mo	€600/yr (€50/mo)	10,000	20
Pro (50k)	n8n Cloud	€145/mo	€1,450/yr (about €121/mo)	50,000	50
Business	Self-hosted only	from €800/mo	from €8,000/yr (about €667/mo)	from 40,000	Set by your own setup
Start-up (discounted Business)	Self-hosted only	see Start-up page	€333/mo billed annually	40,000	Set by your own setup
Enterprise	Cloud or self-hosted	Contact sales	Contact sales	Custom	200+

What else changes between tiers:

Starter: 1 shared project, agents, MCP and AI nodes, 1,600 Assistant credits a month, AI models without your own API keys (paid from Gateway credits, explained below), and forum support.
Pro: everything in Starter plus 3 shared projects, 7 days of insights, admin roles, metric-based evaluations, and 5 days of workflow history. It includes 4,000 Assistant credits on the 10k tier and 9,600 on the 50k tier.
Business: everything in Pro plus 6 shared projects, SSO and LDAP, 30 days of insights, multiple environments, queue mode for scaling, and version control using Git. n8n describes it as "for companies with < 100 employees." Higher Business tiers on monthly billing run from 50,000 executions (€985/month) up to 500,000 (€5,245/month).
Enterprise: unlimited shared projects, 365 days of insights, an external secret store, log streaming, extended data retention, and dedicated support with an SLA.

Every plan includes unlimited users, unlimited workflows, and every integration. You don't pay per seat.

What counts as an execution

This is the number that decides your bill, so it's worth getting exactly right. According to n8n's docs, only production executions count toward your quota. Those are runs that start automatically from a trigger, a schedule, or polling, after you publish the workflow.

How each trigger type counts:

Schedule Trigger: one execution every time it fires, whatever the outcome.
Polling triggers (for example, the Google Drive Trigger): one execution only when new data is found. Polls that come back empty are free.
Webhook Trigger: one execution for every inbound request that activates it, including requests with an empty body like {}. Malformed requests that fail before the workflow starts don't count.

These don't count:

Manual runs from the editor while you build and test
Sub-workflow runs (only the parent execution counts)
Runs of a workflow set as your error workflow
Empty polls and rejected webhook requests

How many executions do you actually need?

For scheduled workflows, the math is simple: count how often the schedule fires in a month. Here is what common schedules use for one workflow in a 30- or 31-day month:

Schedule	Executions / month	Fits Starter (2,500)?	Fits Pro 10k?
Once a day	30–31	Yes	Yes
Every hour	720–744	Yes	Yes
Every 30 minutes	1,440–1,488	Yes, with little room left	Yes
Every 15 minutes	2,880–2,976	No	Yes
Every 10 minutes	4,320–4,464	No	Yes
Every 5 minutes	8,640–8,928	No	Yes, barely

n8n's own pricing FAQ gives the same ballpark: a workflow every 5 minutes uses about 8,600–8,900 executions a month.

For webhooks, your executions equal the number of events that hit the webhook. For a chatbot, n8n suggests estimating weekly conversations and multiplying by the average number of messages per conversation.

A quick worked example. Say you run three workflows: a lead-capture webhook that gets about 20 form submissions a day, an hourly CRM sync, and a daily summary email. That's roughly 620 + 744 + 31 = about 1,395 executions in a 31-day month. Starter covers it. Add one more workflow that checks an API every 15 minutes, and you're at about 4,370, which needs Pro.

What 1,000 executions really cost

Dividing each Cloud plan's price by its allowance shows how quickly the per-run cost drops as you move up. This is our own arithmetic from n8n's listed prices:

Plan	Monthly billing, per 1,000 executions	Annual billing, per 1,000 executions
Starter (2,500)	€9.60	€8.00
Pro (10,000)	€6.00	€5.00
Pro (50,000)	€2.90	about €2.42
Extra executions on Pro 50k (monthly)	€2.00 + tax	not self-serve

The catch is that you only get these rates if you actually use the allowance. A Pro 50k plan running 12,000 executions a month is a much worse deal per run than Pro 10k. Size the plan to your real usage first.

What happens when you hit the limit

On n8n Cloud, this is the part to plan for, because runs over the limit aren't queued. n8n's Help Center explains that once you reach your monthly limit:

Your triggers keep firing, but each new execution fails immediately with "Execution limit reached. Consider upgrading your plan."
There's no backlog. Executions that failed while you were over the limit don't re-run on their own after you upgrade or the month resets. You'd have to retry them manually.
Manual editor runs, chat executions, and sub-workflow runs on a paid plan aren't affected.

A few more rules worth knowing:

Quotas reset on the 1st of each calendar month, not on your billing date.
Annual billing doesn't give you a yearly pool. You get the same monthly allowance either way. Annual only lowers the price.
Upgrading takes effect immediately but doesn't reset what you've already used this month.
n8n offers a one-time courtesy reset of a month's count on the Trial, Starter, and Pro 10k plans, once per customer.
On Pro 50k with monthly billing, you can add executions yourself at €2 (or $2 or £2) plus tax per 1,000, up to 500,000 a month. Annual Pro 50k customers have to email n8n support instead.

Don't confuse the monthly limit with the concurrency limit (5, 20, or 50 runs at the same time on Cloud). When you hit concurrency, extra runs wait in a queue and then run in order. Nothing fails.

Self-hosted Business and Enterprise work differently. If you go over your quota there, n8n says your workflows keep running and overages may be invoiced 45 days later if you don't move up a tier. Business overage is listed at €4,000 per extra bucket of 300,000 executions. The self-hosted license key pings n8n's license server daily, and that ping includes your production execution count.

Three kinds of credits (and which ones cost extra)

n8n now has separate balances that are easy to mix up:

Assistant credits pay for n8n's built-in Assistant (in preview), which helps you build and edit workflows. Each Cloud plan includes a monthly allowance (1,600 on Starter, 4,000 or 9,600 on Pro). They refresh monthly and don't roll over, and you can buy more. On self-hosted n8n you bring your own model key, so no credit limit applies.
Gateway credits let workflow nodes call AI models (such as OpenAI, Anthropic, and Google Gemini) and some tool services without your own API keys. They're a prepaid balance billed per request, separate from your subscription, and only on Cloud Starter and Pro. Top-ups expire 12 months after purchase and aren't refundable except where the law requires it.
Your own API keys still work on every plan. If you already pay OpenAI or Anthropic directly, you don't need Gateway credits at all.

Topping up one kind of credit never adds to the other.

Cloud or self-hosted?

Choose n8n Cloud if you don't want to run a server. n8n handles hosting, and Cloud data is stored in the EU, on servers in Frankfurt, Germany. Starter or Pro is where most solo builders and small teams land.

Choose Community Edition if you're comfortable running Docker and maintaining a server. You pay for hosting, but n8n doesn't charge per execution. You also take on updates, backups, and security yourself, and you give up Cloud-only extras like Gateway credits.

Choose self-hosted Business only if you need SSO, Git-based version control, multiple environments, or queue mode, and you're already running your own infrastructure. If your company has fewer than 20 employees and less than €5 million in total funding, check the Start-up plan first: it's Business at 50% off (€333/month billed annually, 40,000 executions), available for one year or longer while you still qualify.

Which plan should you pick?

You are…	Pick	Why
Just trying n8n	Free Cloud trial	1,000 executions, Pro-level features, no credit card
Running a few hourly or daily automations	Starter	2,500 executions covers several low-frequency workflows
Polling or syncing every 5–15 minutes, or handling real webhook traffic	Pro 10k	A single 15-minute schedule already exceeds Starter
Running 10,000–50,000 executions a month	Pro 50k	Lowest Cloud cost per run, plus self-serve add-ons on monthly billing
A developer happy to maintain a server	Community Edition	Free software. You pay only for your hosting
An early-stage startup that needs SSO and Git	Start-up plan	Business features at half price, self-hosted
A larger company with compliance needs	Enterprise	Custom executions, 200+ concurrency, SLA support

Ways to use fewer executions

These all follow from n8n's own counting rules:

Prefer polling trigger nodes over a Schedule Trigger plus a check. Polling triggers only count when they find new data, while a schedule counts every time it fires.
Process items in batches. One execution can handle any amount of data, so a run that processes 200 rows at once costs the same as one that processes a single row.
Split big workflows into sub-workflows. Only the parent execution counts.
Test with manual runs. Editor runs are free, so keep workflows unpublished until they're ready.
Filter webhook events at the source. Every request that activates the trigger counts, even an empty one, so send n8n only the events you need.
Slow down schedules that don't need speed. Going from every 15 minutes to every 30 roughly halves that workflow's usage.

Free trial details

The n8n Cloud free trial includes Pro-level features such as Global Variables, Insights with 14-day history, 5-day workflow history, unlimited evaluations, and Assistant. Its limits are 1,000 executions, 5 concurrent executions, a 180-second execution timeout, and 1 team project, plus 800 Assistant credits and 100 AI credits. No credit card is required. A Business plan trial gives you full Business features for 14 days, but it does need a card.

See the n8n directory entry for a quick plan summary, or browse more tools in AI Automation & Productivity.

FAQ

Is n8n free? The self-hosted Community Edition is free. n8n Cloud is paid after the free trial, starting at €24/month (€20/month billed annually).

Does n8n charge per user? No. All plans include unlimited users and workflows.

Do failed or test runs use my quota? Manual test runs from the editor don't count. For live runs, n8n's own sources differ: the docs say a Schedule Trigger counts every time it fires "regardless of outcome," while a Help Center article says failed executions don't count. Plan as if every scheduled run counts, and check your Insights dashboard to see what's actually being counted.

Can I get a refund? For the Business plan, n8n's pricing FAQ says it doesn't offer pro-rata refunds. A monthly Business subscription can be cancelled at any time and ends at the close of the billing cycle. An annual one ends at the end of the term.

Can I pay by invoice? Yes, for the annual Business plan and Enterprise, by invoice and wire transfer.

Sources

n8n pricing page and pricing FAQ: https://n8n.io/pricing/
n8n Start-up plan: https://n8n.io/pricing/startup-plan/
n8n Docs, Understand executions: https://docs.n8n.io/build/understand-workflows/understand-executions
n8n Docs, Understand concurrency: https://docs.n8n.io/manage-cloud/concurrency/
n8n Docs, Gateway credits: https://docs.n8n.io/deploy/use-n8n-cloud/gateway-credits
n8n Help Center, How can I add more workflow executions?: https://support.n8n.io/article/can-i-add-more-executions
n8n Help Center, Can you reset my executions?: https://support.n8n.io/article/can-you-reset-my-executions