n8n Pricing Explained: Self-Hosted vs Cloud
This page may contain affiliate links. If you buy through them we may earn a commission at no extra cost to you.
Quick answer: self-hosted n8n is free indefinitely, your only real cost is a small VPS running $6 to $12 a month. If you'd rather not manage a server, n8n Cloud starts around $24/month for a Starter tier with capped active workflows and monthly executions, rising through higher tiers that add more execution volume and admin controls like SSO. This guide breaks down what you're actually paying for on each path, and the costs people forget to budget for on both.
Self-hosted: free in dollars, not in time
n8n's Community Edition is open source and free to run indefinitely under its Sustainable Use License, which permits normal internal business use but restricts reselling it as a competing hosted service. For a small team, that means the honest dollar cost is whatever VPS you deploy it on, typically $6 to $12 a month for a box capable of running a handful of production workflows comfortably. There's no execution cap, no per-task fee, and no plan upgrade waiting for you as usage grows; once the server is paid for, running one more workflow costs nothing extra.
The real cost self-hosting asks for is time, not money. You own updates, database backups, and uptime monitoring yourself, and a misconfigured environment variable (the webhook URL is the classic one) can silently break triggered workflows until you catch it. Our guide to self-hosting n8n walks through the production setup we use, including the specific settings that trip up a first deployment.
n8n Cloud: what each tier actually gets you
If nobody on your team wants to own a server, n8n Cloud removes that maintenance burden entirely, at a published, self-serve price. The Starter tier begins around $24 a month and includes a capped number of active workflows and monthly executions, enough for a small team running a handful of real automations. Higher tiers raise both caps and add features larger teams need: more generous execution volume, and further up, admin controls like SSO and more granular user management.
| Tier | Roughly | Active workflows | Executions | Best for |
|---|---|---|---|---|
| Starter | From ~$24/month | Capped, small number | Capped monthly allowance | Small team, a handful of real workflows |
| Pro | Higher, published tier | Higher cap | Higher monthly allowance | Growing team running more automations |
| Enterprise | Custom, contact sales | Custom | Custom | Larger orgs needing SSO, dedicated support, and governance |
Exact prices and caps shift as n8n evolves the product, so check n8n's own pricing page for current numbers before budgeting rather than relying on any figure here as fixed.
Executions, not tasks or operations
The unit that matters for n8n Cloud billing is the execution, one full run of a workflow, start to finish, regardless of how many nodes that workflow contains. This is a genuinely different model than the two platforms n8n gets compared to most often. Zapier counts every completed action inside a Zap as a separate task, so a workflow with five actions burns five times the allowance of a single-action one. Make counts operations, which scale even more granularly, often per row processed inside a loop. n8n's execution-based model means a complex, ten-node workflow and a simple two-node one cost the same per run, as long as each only runs once. Our n8n review covers how this plays out across real workflows we built on both self-hosted and cloud.
That said, a workflow that runs very frequently, firing on every webhook event from a high-volume source, still adds up fast under a cloud plan's execution cap, even though each individual run is "free" relative to Zapier's per-action meter. High-frequency, high-volume automation is exactly the case where self-hosting's flat cost structure pulls furthest ahead of any cloud plan's cap.
The costs people forget to budget for
Self-hosting's time cost is real and recurring. Every dollar saved on hosting fees is time spent somewhere else, applying updates, rotating credentials, and occasionally debugging why a webhook stopped firing after a server restart. If that time has a real cost to your business, it belongs in the comparison against a cloud plan's price, not just the VPS bill.
LLM API costs are separate from n8n's own pricing. AI Agent and LangChain nodes ship at no extra charge on either self-hosted or cloud n8n, but the moment an agent calls an LLM, you're billed by whichever provider you connected, OpenAI, Anthropic, or otherwise, based on your own token usage. Budget that separately; it's not part of n8n's pricing at all.
Cloud overage and cap behavior matter before you commit. What happens when you exceed a tier's execution cap, whether workflows pause, throttle, or the plan simply requires an upgrade, is worth confirming directly on n8n's current pricing page before you build around a specific tier, since this detail affects a busy month more than the headline price does.
Which path actually makes sense
Self-host if your team is comfortable owning a small server and wants the lowest possible dollar cost at any real volume; that's where n8n's pricing model pulls furthest ahead of task- or operation-billed competitors like Zapier and Make. Pay for n8n Cloud if you want zero infrastructure to manage and your workflow volume comfortably fits a published tier, since the price buys you maintenance-free running, not just raw hosting. For the full three-way cost comparison against Zapier and Make, our Zapier vs Make vs n8n guide walks through pricing models side by side, and our Zapier pricing explained breakdown covers the per-task model n8n's execution billing is most often compared against.
Bottom line
n8n's pricing comes down to a real choice, not a hidden gotcha: pay nothing but your own time by self-hosting, or pay a published, predictable cloud fee and own none of the infrastructure. Neither path bills you per action the way Zapier does, which is the detail that makes n8n meaningfully cheaper than task-based platforms once your workflows run at real volume. Start wherever matches your team's appetite for owning a server, and revisit the decision once your execution volume or active-workflow count is pushing against a cloud tier's cap. Check n8n's own pricing page directly before you budget — tier names and execution caps shift as the product evolves faster than any third-party breakdown can track.
Frequently asked questions
Is self-hosted n8n really free?
Yes, for normal internal business use. n8n's Community Edition is free indefinitely under its Sustainable Use License, which restricts reselling it as a competing hosted service but places no limit on running it for your own company's workflows. Your actual out-of-pocket cost is the server it runs on, typically a $6 to $12 a month VPS, plus the time you spend on updates and backups.
What does n8n Cloud actually bill you for?
Workflow executions, not individual steps or actions within a workflow. A single workflow run counts as one execution regardless of how many nodes it contains, which is a meaningfully different model from Zapier's per-task billing or Make's per-operation billing, both of which charge more as a workflow does more inside a single run.
Should I self-host n8n or pay for the cloud plan?
Self-host if you or someone on your team is comfortable owning a small server, updates, and backups, since the dollar cost is lower almost indefinitely. Pay for n8n Cloud if you'd rather not manage infrastructure at all and the plan's execution cap comfortably covers your workflow volume; the cloud price buys you zero maintenance, not just hosting.
Does n8n's pricing change if I add AI agent workflows?
Not directly through n8n's own billing, AI Agent and LangChain nodes are included at no extra platform cost on both self-hosted and cloud. Your real added cost is whatever LLM API you connect the agent to, which bills separately based on your own usage with that provider.