TL;DR — n8n has two prices, and picking the right one changes your bill by an order of magnitude. Self-hosted is free with unlimited executions — you pay only for a server (~€5–20/month). n8n Cloud is managed and billed per execution (one workflow run, regardless of steps): Starter €24/mo (2,500 executions), Pro €60/mo (10,000), Business €800/mo (40,000, with SSO), Enterprise on quote. There's a 14-day Cloud trial (no card) and ~17% off for annual. This guide models what you'll actually pay on each path — and where self-hosting quietly wins.
The one rule that explains n8n pricing
n8n bills on executions, and an execution is a whole workflow run — trigger to finish — no matter how many nodes, branches, loops or AI steps it contains[1]. That's the opposite of per-task tools, where every step adds to the meter. A twenty-node n8n workflow that runs once is one execution. Since April 2026, every plan also includes unlimited active workflows, so on Cloud the only number that matters is how many times your workflows run per month. On self-hosted, even that is free.
The two paths
Self-hosted (Community Edition) — free. You run n8n on your own infrastructure (Docker is the usual route) with unlimited executions and no licence cost. Your only spend is the server: a small VPS at roughly €5–20/month comfortably runs a lot of workflows. The trade is operations — you own updates, backups and security — but the price ceiling is essentially the hardware.
n8n Cloud — managed, per execution. n8n hosts everything; you pay for a plan sized to your monthly executions. This buys back all the operations work, which is worth real money if you don't have the technical capacity to run a server well.
n8n Cloud plans, decoded
| Plan | Price (monthly) | Executions/mo | Notable |
|---|---|---|---|
| Starter | €24 | 2,500 | 14-day free trial, no card |
| Pro | €60 | 10,000 | More concurrency, admin roles |
| Business | €800 | 40,000 | SSO, advanced permissions |
| Enterprise | quote | unlimited | Dedicated infra, SLA, governance |
Annual billing takes roughly 17% off each monthly figure. The jump from Pro (€60) to Business (€800) is large because Business is really about governance — SSO and permissions for bigger orgs — not just more executions. If you need 40,000 executions but not SSO, that gap is the moment to seriously price self-hosting.
What self-hosting actually costs at scale
"Free" self-hosting has a real cost curve once you grow, and it's honest to name it. For modest volumes, a single small server running n8n plus a Postgres database is all you need. But past roughly ten simultaneous executions, you'll move to queue mode — n8n's horizontal-scaling setup that spreads executions across worker processes[2]. Queue mode is where the infrastructure bill appears: it wants a PostgreSQL database, a Redis instance, and external storage (like S3) for binary data[2]. None of that is expensive individually, but it's more than one tiny VPS, and standing it up is a migration project rather than a checkbox. Even so, for a technical team the all-in infrastructure cost of a scaled self-hosted n8n is typically a fraction of the equivalent Cloud plan — you're trading money for engineering time.
How executions add up — and what inflates them (Cloud)
On Cloud, forecast executions by counting workflow runs, then watch three multipliers:
- Sub-workflows count separately. A main workflow that calls a sub-workflow per item turns one logical job into two-plus executions. Clean modular design can quietly double your meter — consolidate on Cloud, or self-host where it's free.
- Schedule frequency. A workflow polling every minute is ~43,000 executions a month on its own. Match the interval to real need.
- Error retries. A workflow set to retry a flaky step re-runs and re-counts. Useful, but budget for it.
None of these applies to self-hosted, where executions are unlimited — which is precisely why high-volume teams gravitate there.
Worked examples
A lean team, managed. A dozen workflows running a few thousand times a month sits inside Starter's 2,500 or Pro's 10,000. Real bill: €24–60/month on Cloud, or ~€10/month self-hosted if you'd rather run the server.
A high-volume shop. Say 100,000+ executions a month. On Cloud that points at Business (€800) or Enterprise; self-hosted, the same volume is still free of execution charges — you scale the server (queue mode with Postgres + Redis) for perhaps €40–100/month of infrastructure. The crossover is stark: at high volume, self-hosting can be 5–10× cheaper, in exchange for owning the ops.
An AI-heavy workflow. Remember the model is billed separately by your provider (you bring your own key), so an agent that calls an LLM many times per run adds a provider cost on top of the n8n execution. Budget both meters when an AI-Agent loop is involved.
12-month total cost, illustrative
| Profile | Path | ~Monthly | ~12-month |
|---|---|---|---|
| Technical solo/team, modest volume | Self-hosted (1 small VPS) | €5–20 | ~€60–240 |
| Non-technical small team | Starter Cloud (annual) | ~€20 | ~€240 |
| Growing volume, managed | Pro Cloud (annual) | ~€50 | ~€600 |
| Need SSO / governance | Business Cloud | €800 | ~€9,600 |
| High volume, technical | Self-hosted, queue mode | €40–100 (infra) | ~€480–1,200 |
The two rows that tell the story are the last two: the same high volume that costs ~€9,600/year on Business Cloud can run for roughly €480–1,200/year self-hosted. That's the trade in numbers — thousands of euros a year, in exchange for the engineering time to run Postgres, Redis and a couple of workers.
The other bill: your AI model provider
One line item catches AI-heavy teams off guard, and it's worth stating plainly. Because n8n uses your own model and key, the language-model usage is billed by that provider (OpenAI, Anthropic, or your own hosted model) — completely separately from n8n. On the plus side, you see and control that cost directly, and self-hosting an open-weight model can drive it to near-zero. But an AI-Agent workflow that calls a model many times per run can make the model bill larger than your n8n bill. When you budget an AI workflow, budget two meters: n8n executions and your model provider's tokens.
Which plan fits you
| You are… | Choose… |
|---|---|
| Technical, want lowest cost + data ownership | Self-hosted Community Edition (free) |
| Non-technical, small/modest volume | Starter Cloud (€24) after the trial |
| Growing execution volume, want it managed | Pro Cloud (€60) |
| A larger org needing SSO/permissions | Business Cloud (€800) |
| Regulated or very high volume | Enterprise, or self-host at scale |
The crossover: when self-hosting wins
The decision comes down to two questions: do you have the technical capacity to run a server well, and how high is your volume? If you're technical and running meaningful volume, self-hosting almost always wins on cost, and it adds data ownership for free. If you're not technical, or your volume is modest and your time is better spent elsewhere, Cloud's convenience is worth the subscription. The nice part: because workflows export as JSON, you can start on Cloud and migrate to self-hosted later without losing your work — the escape hatch is always open.
No per-seat pricing — a quiet saving
One thing n8n's pricing page doesn't shout, but your finance team will appreciate: there's no per-user charge. Cloud plans bill on executions and self-hosting bills on nothing, so adding teammates to build and manage workflows costs you nothing extra either way. For tools that charge per editor or per seat, a growing automation team is a growing bill; on n8n, headcount and cost are decoupled. It's a meaningful structural saving for any team where more than one or two people touch the automations.
Trials, discounts, and avoiding surprises
Start free either way: self-host the Community Edition (free, unlimited executions), or take the 14-day Cloud trial (Starter/Pro, no credit card). Commit annually only once n8n is clearly part of your stack — that locks in the ~17% saving. And before you scale on Cloud, forecast executions honestly (runs × frequency, plus sub-workflow calls); before you scale self-hosted, budget for the Postgres/Redis/S3 that queue mode needs. Either surprise is entirely avoidable with five minutes of modeling.
The bottom line
n8n is one of the best-value automation platforms in the market — but the value depends on choosing the right path. If you're technical, self-hosting gives you unlimited executions for the price of a small server and full data ownership; if you're not, Cloud's per-execution model (starting at €24/month) is fair and predictable, and far cheaper than per-task tools for complex workflows. Model your executions, weigh your capacity to run infrastructure, and you'll land on the right one. See the full analysis in our n8n review, or start with the tutorial.