n8n Pricing: Every Plan, Cost & the Free Self-Hosted Option
If you’re comparing workflow automation tools in 2026, n8n keeps coming up for one reason: it prices completely differently from Zapier and Make. Instead of charging you for every single step your automation runs, n8n charges per full workflow execution — one complete run from trigger to finish counts as one execution, whether it has 2 steps or 40. That single design choice can cut costs dramatically for complex, AI-heavy workflows.
But n8n’s pricing has real quirks too: a steep cliff between Pro and Business, euro-denominated plans, and a free self-hosted edition with a “fair-code” license that isn’t quite open source. Here’s the full breakdown of every n8n plan in 2026, what each one actually gets you, and how to decide between cloud and self-hosted.
n8n Pricing at a Glance
| Plan | Price (billed annually) | Executions / month | Best for |
|---|---|---|---|
| Community (self-hosted) | Free | Unlimited | Developers, privacy-focused teams |
| Starter (cloud) | €20/mo | 2,500 | Solopreneurs, light automations |
| Pro (cloud) | €50/mo | 10,000 | Growing teams, agencies |
| Business | €667/mo | 40,000 | Enterprises with DevOps capacity |
| Enterprise | Custom | Custom | Large orgs, compliance needs |
All prices above are the annual-billing rates published on n8n’s official pricing page. Paying monthly costs meaningfully more — the annual toggle advertises roughly 17% savings.
How n8n’s Execution-Based Pricing Actually Works
This is the most important thing to understand before comparing n8n to anything else. Zapier charges per “task” (each step), and Make charges per “operation” (each step). n8n charges per execution: one complete workflow run.
A 10-step workflow that runs 1,000 times a month costs you 1,000 executions on n8n — but 10,000 tasks on Zapier. For simple two-step zaps the difference barely matters. For 30-step AI pipelines with branching, loops, and API enrichment, n8n’s model is structurally far cheaper. If your workflows are long and complex, this is n8n’s killer pricing advantage.
n8n Cloud Plans Explained
Starter — €20/month
The entry cloud tier, aimed at individuals and very small projects:
- 2,500 workflow executions per month
- 5 active workflows, 5 concurrent executions
- 1 shared project, 1 user seat structure (unlimited users is a higher-tier story — check current terms)
- AI workflow builder credits included
- Cloud-hosted: no server to manage, automatic updates
Starter is a reasonable on-ramp if you’re automating personal productivity or running a handful of business workflows. The 2,500-execution cap goes fast once you add polling triggers (e.g., checking an inbox every 5 minutes burns ~8,600 executions a month by itself), so watch trigger frequency.
Pro — €50/month
The sweet spot for most small teams and agencies:
- 10,000 workflow executions per month
- 20 concurrent executions
- 3 shared projects
- Workflow history, global variables, admin roles, execution search
- More AI credits and priority support
Pro is where n8n starts to seriously undercut Zapier for equivalent volume. A team running multi-step lead routing, content pipelines, or AI agent workflows will usually find Pro’s 10,000 executions go 3–5x further than a comparably priced Zapier tier.
Business — €667/month
Here’s the cliff: Business costs more than 13x Pro. What you get:
- 40,000 executions per month
- 6 shared projects
- SSO, SAML, and LDAP authentication
- Git-based version control and multiple environments
- 30-day insights retention
- Self-hosted deployment on the same license
Business is explicitly aimed at organizations that already have DevOps capacity and just need the license plus enterprise controls. Startups with fewer than 20 employees can get 50% off Business pricing — a significant discount worth asking about.
Enterprise — Custom
For large organizations with compliance and scale requirements:
- Custom execution quotas, 200+ concurrent executions
- Unlimited shared projects
- Log streaming, external secret stores, 365-day execution log retention
- SSO, audit logs, dedicated support, SLA options
- Cloud or self-hosted
The Free Option: Self-Hosted Community Edition
n8n’s Community Edition is free to self-host with unlimited executions. You pay only for the server — commonly €5–20/month on a small VPS from providers like Hetzner or DigitalOcean, plus a free domain with Let’s Encrypt SSL. Total realistic cost: under €20/month for unlimited automation.
The trade-offs are real, though:
- You handle everything: updates, backups, monitoring, security patches (1–2 hours a month for a simple setup)
- You need basic DevOps knowledge: Docker is the standard deployment path
- The license is “fair-code,” not open source: n8n uses the Sustainable Use License. You can use it freely for internal business purposes, but you cannot sell a product built on n8n or charge users for access to it. Read the license if you plan anything commercial.
For a technical founder or a team with one DevOps-capable member, self-hosted n8n is arguably the best value in all of workflow automation.
n8n vs Zapier vs Make: Real Cost Comparison
Take a realistic scenario: a 10-step workflow (lead capture → enrichment → AI scoring → CRM update → Slack alerts → follow-up sequences) running 10,000 times a month.
| Platform | What you’re billed | Approx. monthly cost |
|---|---|---|
| n8n (cloud Pro) | 10,000 executions | ~€50 |
| n8n (self-hosted) | Server only | ~€5–20 |
| Zapier | 100,000 tasks | ~$300+ |
| Make | ~100,000 operations | ~$65+ |
n8n wins this comparison by a wide margin, and the gap widens as workflows get longer. Zapier’s advantage is its 7,000+ integrations and non-technical ease of use; Make sits in the middle on both price and power. n8n’s 400+ integrations cover the major tools, plus you get full JavaScript/TypeScript code nodes and custom npm nodes for anything missing.
Which n8n Plan Should You Choose?
Choose Community (self-hosted) if: you’re technical, data must stay on your infrastructure (GDPR, healthcare, finance), you need unlimited executions, or you want to connect to internal systems behind a firewall.
Choose Starter if: you’re an individual automating personal workflows and you want zero server management.
Choose Pro if: you’re a small team or agency running real business volume — this is the best value cloud tier for most buyers.
Choose Business if: you need SSO, Git version control, and multiple environments, and you have the DevOps capacity to self-host (or want n8n-hosted enterprise controls).
Choose Enterprise if: you’re a large organization with compliance requirements (SOC 2, HIPAA-style controls), audit needs, and high concurrency demands.
If you’re also evaluating AI assistant pricing alongside automation tools, our guides to Perplexity’s pricing plans and Grok’s pricing tiers break down what the major AI platforms charge in 2026 — useful context when budgeting an AI-powered workflow stack. And if your automation involves AI-generated audio, check Suno’s pricing too.
n8n’s AI Features and What They Cost
In 2026, a big reason people choose n8n is AI orchestration — chaining LLM calls, building agents, and wiring AI into business workflows. n8n ships with 70+ AI/LangChain nodes covering OpenAI, Anthropic, Google, and local models.
The big platforms are all building agent layers too — see our breakdown of what Microsoft Agent 365 is and what it costs.
On cloud plans, AI usage is metered partly through AI credits bundled with each tier (Starter includes a modest allowance; Pro includes more). Heavy AI workloads can burn through these, so check your plan’s AI credit allocation if your workflows call LLMs thousands of times a day. The models themselves bill separately through your own API keys — n8n doesn’t resell tokens, you bring your own OpenAI/Anthropic/Google credentials.
Self-hosted users face no n8n-side AI metering at all: your only AI cost is whatever your model provider charges. This is another reason AI-heavy teams gravitate toward self-hosting.
Hidden Costs the Pricing Page Doesn’t Spell Out
n8n’s pricing is refreshingly simple, but a few costs catch new users off guard:
- Polling triggers eat executions. A workflow that checks Gmail every minute consumes ~43,000 executions a month before doing any real work. Use webhooks instead of polling wherever possible — they’re instant and far cheaper.
- Monthly billing premium. The ~17% annual discount means monthly payers leave real money on the table. If you’re confident after a trial month, switch to annual.
- AI credits on cloud plans. If your workflows are AI-heavy, model API costs (OpenAI, Anthropic) sit on top of your n8n subscription. Budget both.
- Self-hosting isn’t literally free. A VPS, backups, and your maintenance time are real costs. For most small setups it’s €5–20/month plus an hour or two of upkeep — still a fraction of any cloud automation bill at volume.
- The Business tier cliff. Going from Pro (€50) to Business (€667) is a 13x jump, mostly for SSO, Git, and environments. If you need just one of those features, evaluate whether the jump is worth it or whether a workaround exists.
Three Real-World Pricing Scenarios
The Freelancer: €20/month (or ~€8 self-hosted)
Maya runs a solo content agency. She has 6 workflows: client onboarding, invoice reminders, social scheduling, lead capture, testimonial collection, and a weekly report digest. Total: ~1,800 executions/month. Starter at €20/month covers her with headroom — or she could self-host on a €6 VPS and pay almost nothing. She chose Starter for zero maintenance.
The Agency: €50/month
A 12-person marketing agency runs 40 workflows: multi-step lead routing with AI scoring, client reporting pipelines, CRM syncs, and content production chains averaging 15 nodes each. Volume: ~8,000 executions/month. On Zapier, those 15-node workflows would bill ~120,000 tasks — well into $300+/month territory. On n8n Pro: €50 flat. Annual savings vs. Zapier: roughly $3,000+.
The SaaS Company: Business tier
A 60-person SaaS company needs SSO for security compliance, Git version control for its workflow codebase, and separate dev/staging/prod environments. They self-host on their own Kubernetes cluster and pay for the Business license (€667/month, or ~€333 with the startup discount while eligible). Compared to building equivalent automation infrastructure in-house, it’s a clear win.
How to Keep Your n8n Bill Low
- Prefer webhooks over polling triggers. This single change can cut execution usage by 80%+ for event-driven workflows.
- Batch aggressively. Process 100 rows in one execution instead of triggering 100 executions.
- Start on Pro monthly, then commit annually once you know your volume.
- Migrate to self-hosted when the math flips. Rule of thumb: if you’re consistently near Pro’s 10,000-execution ceiling and have any technical capacity, a €10 VPS pays for itself immediately.
- Claim the startup discount. Under 20 employees? Ask about the 50% Business discount before paying full price.
What Happens When You Exceed Your Execution Limit?
On cloud plans, blowing past your monthly execution quota doesn’t silently rack up overage fees the way some platforms do — but it does stop or throttle your workflows until the next cycle (or until you upgrade). That’s why monitoring matters: n8n’s dashboard shows execution usage, and you should set up alerts at 70–80% of your quota.
If you hit the ceiling regularly, you have three options: upgrade to the next tier, optimize trigger frequency and batching (see above), or move to self-hosted Community where executions are unlimited. Many teams do a hybrid: keep prototyping on n8n Cloud, run production volume self-hosted.
A Brief History of n8n Pricing
n8n’s pricing has evolved with its audience. Early on, it was almost entirely a self-hosted, developer-loved tool with a simple paid tier for teams wanting extras. As AI workflows exploded in 2024–2025, n8n leaned into cloud plans with execution-based pricing — deliberately positioned against Zapier’s per-task model — and added AI builder credits, shared projects, and enterprise controls like SSO and Git integration.
The fair-code Sustainable Use License has been the constant through it all: n8n’s bet is that giving the core product away (with unlimited self-hosted executions) grows the community, while teams that want managed hosting, collaboration, and compliance pay for cloud or Business/Enterprise licenses. In 2026, with 400+ integrations and deep AI node support, that bet looks like it’s paying off.
Frequently Asked Questions
Is n8n really free?
The self-hosted Community Edition is free forever with unlimited executions — you only pay for your own server (typically €5–20/month). The cloud plans start at €20/month. So yes, n8n can genuinely be free if you’re willing to self-host.
What counts as an “execution” in n8n?
One complete run of a workflow, from trigger to final node, counts as one execution — regardless of how many steps (nodes) the workflow contains. A 40-node AI pipeline and a 2-node notification both cost exactly one execution per run.
Is n8n open source?
Not exactly. n8n’s code is publicly available under the Sustainable Use License, a “fair-code” license. You can use, modify, and self-host it freely for internal business use, but you can’t offer n8n itself as a paid service or embed it in a commercial product without a commercial arrangement.
How does n8n compare to Zapier on price?
For complex multi-step workflows, n8n is dramatically cheaper because it bills per execution rather than per step. A workload costing ~€50/month on n8n Pro can easily cost $300+/month on Zapier. Zapier’s edge is simplicity and its 7,000+ app integrations.
Can I switch from n8n Cloud to self-hosted later?
Yes. Workflows export as JSON and import cleanly into a self-hosted instance. Many teams start on n8n Cloud for speed and migrate to self-hosted as volume grows — it’s one of the platform’s genuine strengths.
Does n8n offer a startup discount?
Yes — startups with fewer than 20 employees can receive 50% off the Business tier when billed annually. Check n8n’s pricing page or contact their sales team for current terms.
Does n8n charge per user or per seat?
One of n8n’s most unusual pricing decisions: paid cloud plans don’t meter seats the way most SaaS does. You get unlimited users and unlimited workflows on every paid tier — you’re billed on executions, not headcount. For teams, this removes the annoying “add another $20 seat” math that plagues Zapier and Make budgeting.
Can I use n8n for client work as an agency?
Yes — agencies commonly run client automations on n8n Pro or self-hosted instances. The main constraint is the Sustainable Use License: you can’t repackage n8n itself as a paid product or charge users for access to n8n. Running workflows on behalf of clients as a service is standard practice; just make sure each client’s data isolation and your own liability terms are handled properly.
The Bottom Line
n8n’s pricing is built around a simple, powerful idea: charge for workflow runs, not steps. That makes the €50/month Pro plan one of the best values in automation for teams running complex workflows, and the free self-hosted Community Edition is unmatched if you have basic technical chops. The main watch-outs are the steep Pro-to-Business cliff and the fair-code license terms. Start with Pro (or self-hosted Community if you’re technical), measure your actual execution volume for a month, and scale from there.
And revisit the math every six months — as your workflows grow longer and more AI-heavy, n8n’s execution-based model only gets more favorable compared to per-step competitors. For the full official breakdown, see n8n.io/pricing and the main n8n site.
