n8n Pricing 2026: Cloud Plans, Executions and Self-Hosting Cost

n8n Cloud costs $20/mo on Starter for 2,500 workflow executions and $50/mo on Pro for 10,000, both billed annually, or $24 and $60 on monthly billing. The Business plan is $800/mo on annual billing for 40,000 executions and runs on your own servers only. The Community edition is free to self-host, so the second half of this page prices the same workload on a $12 and a $24 server.

n8n bills executions, not steps or users. One execution is one run of a whole workflow, however many nodes it has, and every plan includes unlimited users and workflows. That is why a chatbot or an agent with 12 steps costs the same per run as a two-node webhook, and why the self-hosted edition, which has no execution meter at all, is the price to beat.

Prices were read on n8n's pricing page and its US price feed on 26 October 2026, the server prices on DigitalOcean's Droplet pricing the same day. Whether n8n is the right tool is the n8n review; the step-by-step install is how to self-host n8n; the Zapier cost comparison at 10,000 runs is n8n vs Zapier.

Toolradar data: Toolradar, the software directory we run, tracks 1,031 workflow automation tools as of October 26, 2026. 36% of them offer a free tier, and the median cheapest paid plan is $39/mo across the 238 tools that publish one, so n8n's Starter sits well under the category's typical entry price. The figures are on the workflow automation statistics page and the tool record is n8n on Toolradar, where it ranks #16 most-listed alternative among the category's 1,030 tools. Devshot data: this newsletter reaches 31,000+ developers as of November 8, 2026, with an open rate of 32%.

At a glance

Plan	Price, annual billing	Price, monthly billing	Executions a month	Where it runs	Concurrent executions
Community	$0	$0	Unmetered	Self-hosted	Set by your server
Starter	$20/mo ($240 a year)	$24/mo	2,500	n8n Cloud	5
Pro	$50/mo ($600 a year)	$60/mo	10,000	n8n Cloud	20
Pro 50K	$120/mo ($1,440 a year)	$144/mo	50,000	n8n Cloud	50
Business	$800/mo ($9,600 a year)	$960/mo	40,000	Self-hosted only	Your server
Enterprise	Custom	Custom	Custom	n8n Cloud or self-hosted	200+

Annual plans are charged for the year up front: the Starter figure is $240 charged once, not twelve monthly payments. Monthly billing costs 20% more on every self-serve plan and can be cancelled at the end of any billing cycle. The Business ladder goes on past 40,000, up to 500,000 executions at $5,245/mo on annual billing, with each tier published on the pricing page.

How does n8n pricing work?

An execution is one complete run of a workflow, from trigger to last node. Manual test runs, sub-workflow calls and error workflows do not count, according to n8n's concurrency documentation, and only production executions (those started by a webhook or a trigger node) hit the plan's cap. Each turn of a chat agent is its own execution, so a busy support bot burns the quota faster than a nightly sync.

That meter matters because the competitors count differently. Zapier sells tiers in tasks per month rather than runs, and n8n's own page contrasts its model with tools that "charge per step or user". A multi-step workflow that runs 2,500 times a month is 2,500 executions on n8n and fits Starter, whatever the step count. Zapier's pricing page starts its Professional plan at $19.99/mo on yearly billing for 750 tasks, with a free plan at 100 tasks and Team from $69/mo. The head-to-head at a fixed workload is in n8n vs Zapier.

The other things the plans change are limits, not features: concurrency, how long a single execution may run, and how much history n8n keeps.

Limit	Starter	Pro	Enterprise
Concurrent production executions	5	20 on 10K, 50 on 50K	200+
Maximum execution duration	5 minutes	40 minutes	40 minutes
Saved executions	2,500	25,000	50,000
Execution log retention	7 days	30 days	Unlimited
Shared projects	1	3	Unlimited
Assistant credits a month	1,600	4,000 on 10K, 9,600 on 50K	Not included

Executions beyond the concurrency limit queue and run in order, so the cap slows you down rather than failing runs. The 5-minute execution ceiling on Starter is the limit that bites first for AI workflows: a long agent run or a large file transform that exceeds it stops, and only Pro lifts it to 40 minutes.

What each n8n plan includes

Starter: $20/mo for 2,500 executions

Starter is one shared project, five concurrent executions, every integration, the agent and MCP nodes, and 1,600 Assistant credits a month. Workflow history keeps one day. It fits a solo builder with a handful of scheduled jobs: a workflow that runs every 15 minutes uses about 2,880 executions a month by our arithmetic, already past the cap, so anything polling-heavy belongs on Pro or on your own server.

Pro: $50/mo for 10,000 executions

Pro adds three shared projects, 20 concurrent executions, admin roles, the project editor role, seven days of insights, five days of workflow history, execution search and unlimited metric-based evaluations. The 50,000-execution size at $120/mo on annual billing raises concurrency to 50 and Assistant credits to 9,600. Pro is the plan n8n's own free trial mirrors, and the ceiling of n8n Cloud's self-serve range: past 50,000 executions a month the next hosted step is an Enterprise quote.

Business: $800/mo, self-hosted only

Business is a license key for your own instance, not a hosted plan: the pricing page says a hosted version has a waitlist. For that price on annual billing ($960 monthly) you get 40,000 executions, counted across every instance the key is applied to. The key also unlocks six shared projects, SSO with SAML and LDAP, environments, Git version control, queue mode and 30 days of insights, after a 14-day trial that requires a card. What you are paying for is governance, not capacity: the same server running the free Community edition executes the same workflows with no meter. Overages, when n8n invoices them, are priced in euros at €4,000 per extra block of 300,000 executions.

Start-ups with fewer than 20 employees and under €5M raised qualify for the Start-up plan, 50% off Business for one year, which puts the 40,000-execution tier at $400/mo on annual billing by our arithmetic.

Enterprise: quote only

Enterprise runs on n8n Cloud or self-hosted and adds 200+ concurrent executions, 365 days of insights, external secret stores, log streaming to tools such as Datadog, audit logging, a 2FA requirement, pay by invoice and dedicated support with an SLA. It is the only plan with email or dedicated support: Starter, Pro and Business all point to the community forum.

Community edition: free, with a license to read

The self-hosted Community edition is the full product minus the collaboration and governance layer. Per n8n's editions comparison, it lacks projects, sharing of workflows and credentials beyond their creator, SSO, environments, external secrets, log streaming, Git version control, custom variables and multi-main scaling. Registering the instance with an email adds folders, debug in editor and custom execution data, still free.

The license is n8n's Sustainable Use License, not an OSI open-source license. You may use and modify it for your own internal business purposes, or personal and non-commercial use, and distribute it only free of charge for non-commercial purposes. Running n8n for your company's own automations is fine; selling access to a hosted n8n is not.

Cloud Starter vs a self-hosted server: 2,500 and 10,000 executions

The self-hosted bill is the server, because the software is free. n8n's Docker Compose guide asks for at least 4 GB of RAM and 2 vCPUs, and says the extra headroom is for the Docker-in-Docker sandbox that runs Assistant's generated code. A leaner setup without that sandbox has no documented minimum, so the lower row is our choice of a 2 GiB machine, not an n8n figure.

Workload a month	n8n Cloud, annual billing	Self-hosted, 2 GiB Droplet	Self-hosted, 4 GiB Droplet (the documented spec)
2,500 executions	Starter, $240 a year	$14.40/mo, $172.80 a year	$28.80/mo, $345.60 a year
10,000 executions	Pro, $600 a year	$14.40/mo, $172.80 a year	$28.80/mo, $345.60 a year
50,000 executions	Pro 50K, $1,440 a year	Same server until it saturates	Same server until it saturates

The server rows are our arithmetic on DigitalOcean's list: a Basic Droplet with 2 GiB and 1 vCPU is $12/mo, 4 GiB with 2 vCPUs is $24/mo, and weekly backups add 20% of the Droplet price ($2.40 and $4.80). Daily backups cost 30% instead, which makes the 4 GiB row $31.20/mo. DigitalOcean is the reference because its list is in dollars with no term or renewal catch; Linode, Lightsail and OVHcloud at this size are compared in best cheap VPS.

At 2,500 executions the gap is small: Starter's yearly charge against $173 for the lean server, a $67 difference that disappears after one afternoon of your time on updates and a restore test. At 10,000 executions the gap is $600 against $173, and the self-hosted box has no reason to cost more at 50,000 unless the workflows are heavy. That is the real pattern in n8n's pricing: Cloud scales with runs, a server scales with load, and the two part ways around the Pro tier.

What the server price leaves out is your time: monthly updates, Postgres backups and their restore test, a domain with TLS, the Assistant's sandbox if you want it, and an incident at 3 a.m. with nobody to call. n8n's own decision guide recommends self-hosting for teams that "have infrastructure and technical resources" and n8n Cloud for everyone who does not want to manage any. The install, the Compose file and the update routine are in how to self-host n8n; the Droplet tiers and their bandwidth are in DigitalOcean pricing.

The free trial and the free experiments

n8n Cloud's trial is 14 days with no credit card, per the trial documentation. It runs Pro features (global variables, insights, execution search, five days of workflow history) with Starter's compute, a limit of 1,000 executions, five concurrent executions and a 180-second execution timeout. When the trial ends n8n deletes the instance, and you have 90 days to download your workflows.

There is no free Cloud tier after the trial. The free path is the self-hosted Community edition, which the cloud free tiers comparison places next to the hosting credits that could run it at $0 for a while.

What changed in n8n's pricing

The plan ladder has been stable but the caps inside it have not. In the January 8, 2024 snapshot of the pricing page, Starter carried 2,500 executions and only five active workflows, and Pro offered 10,000 executions with 15 active workflows or 50,000 with 50. Today every plan has unlimited active workflows, which makes Starter far more usable for a builder with many small automations.

The Business plan, with its self-hosted license key and 40,000 executions, appears in the January 2, 2026 snapshot at the same euro prices n8n shows European visitors today, so there has been no self-serve price rise in 2026 so far. What did move during 2026 is the AI allowance: that January page listed 50 "AI Workflow Builder" credits on Starter and 150 on Pro, where today's page gives 1,600 Assistant credits on Starter and up to 9,600 on Pro.

Gotchas and hidden costs

• Annual means prepaid. The $20 and $50 figures are the per-month view of a yearly charge, and n8n's cancellation FAQ says it offers no pro-rata refunds.
• Business is not hosted. Buying it means running your own instance; the Cloud version is a waitlist.
• Overage is documented for Business only. The pricing page says the sales team contacts a Business customer near the quota, and any overage left unresolved is invoiced 45 days later; it prints no overage rate for Starter or Pro.
• Executions are kept, not forever. Starter keeps 2,500 saved executions for 7 days; a debugging session on last week's run needs Pro's 30 days.
• Support is the forum on every plan below Enterprise, including Business.
• Hosted data lives in Frankfurt. n8n Cloud stores data in the EU; a US data-residency requirement means self-hosting.
• Assistant credits do not roll over, and extra credits are sold separately once the monthly allowance is used.

Who should pay for n8n Cloud, and who should not

Pay for Starter or Pro when your time costs more than the plan. A one-person team running under 10,000 executions a month gets backups, updates and uptime for Starter or Pro money, which is less than an hour of maintenance a month at any developer rate. The Pro tier also unlocks the 40-minute execution window that long agent runs need.

Self-host when the workload is heavy or the data cannot leave your network. A Droplet at either size above runs the Community edition with no execution meter, and the self-hosting tools guide covers the deploy layers, such as Coolify, that make the upkeep closer to a managed service. Teams that then need SSO, projects and environments pay for Business, which is a governance purchase, not a capacity one.

Compare against the alternatives when steps are few. A one-step or two-step automation does not benefit from n8n's per-execution model, and Zapier's $19.99 Professional tier or an agent platform such as Dify may cost the same for less setup. Both products have Toolradar profiles, n8n and Zapier, with their plans and alternatives side by side.

Methodology and sources

Plan prices, executions, concurrency, retention and feature lists were read on n8n's pricing page on 26 October 2026, in the US price list that its plan cards render from. Trial terms come from the trial documentation, the execution rules from the concurrency documentation, the Community edition's scope from the editions comparison, the license terms from LICENSE.md and the server requirement from the Docker Compose guide. Start-up plan terms are on n8n's Start-up plan page.

Server prices come from DigitalOcean's Droplet pricing, Zapier's from Zapier's pricing page, both read the same day. The yearly totals, the backup add-ons and the Start-up plan dollar figure are our arithmetic on those printed prices. Past caps come from the dated web.archive.org copies linked above and are used only for the "before" comparison. Toolradar figures are from its workflow automation category as of October 26, 2026.

Louis Corneloup, founder of Toolradar and Dupple, wrote this page. We run no test lab and give no rating. The method is on how we rate, corrections on the editorial policy and the author note on about. Check the pricing page once more before an annual commitment.

FAQ

How much does n8n cost per month in 2026?

n8n Cloud Starter is $20/mo on annual billing or $24/mo on monthly billing for 2,500 executions; Pro is $50/mo annual or $60/mo monthly for 10,000, with a 50,000-execution size at $120/mo annual. Business, for self-hosted instances only, is $800/mo annual or $960/mo monthly for 40,000 executions. Enterprise is a quote.

Is n8n free?

Yes, when you host it yourself. The Community edition is free under the Sustainable Use License, with unlimited executions and almost every feature except projects, sharing beyond the creator, SSO, environments, external secrets, log streaming and Git version control. n8n Cloud has a 14-day free trial with 1,000 executions and no free plan after it.

What counts as an execution in n8n?

One complete run of a workflow, whatever its number of steps or the data volume. Only production executions started by a trigger or webhook count toward the plan; manual tests, sub-workflow calls and error workflows do not. Each turn of a chat agent is a separate execution.

Does n8n charge per user?

No. Every plan, including Starter, includes unlimited users and unlimited workflows. The plans differ by executions, concurrency, retention and the collaboration features (projects, roles, SSO) that come with Pro, Business and Enterprise.

Is self-hosting n8n cheaper than n8n Cloud?

At 2,500 executions a month it is close: $240 a year for Starter against about $173 for a 2 GiB DigitalOcean Droplet with weekly backups, by our arithmetic. At 10,000 executions the server costs the same while Pro is $600 a year, so the saving grows with volume, minus the time you spend on updates and backups.

Can I buy the Business plan on n8n Cloud?

Not at the moment. The Business plan is a license key for self-hosted instances, and n8n's pricing page offers a waitlist for a hosted version. Hosted customers who outgrow Pro's 50,000 executions move to an Enterprise quote.

Cite this: Devshot, "n8n Pricing 2026: Cloud Plans, Executions and Self-Hosting Cost", October 2026.

The daily Devshot brief is where plan changes like these show up before they show up on an invoice. Subscribe at devshot.co.

Some offers on this page may be paid placements or contain affiliate links.