Is n8n Free? n8n Pricing Explained (Self-Hosted vs Cloud)
n8n has a genuinely free self-hosted edition and a paid cloud version. What free actually includes, what self-hosting really costs, and when n8n Cloud is worth paying for.
Gaurav Guha
Co-Founder, SailoLabs
Short answer: yes, n8n is genuinely free to self-host. The Community Edition is source-available, runs on your own server, and has no execution limits, no workflow caps, and no meter running. Thousands of companies run their entire automation stack on it and pay nothing but hosting. Longer answer: "free" hides two real questions. What does self-hosting actually cost you in servers and time? And what are you giving up compared to n8n Cloud, the paid managed version? This guide answers both, with the practical math we use when advising clients on which route to take.
The Two Ways to Run n8n
n8n comes in two flavors, and the pricing conversation is really a choice between them. Self-hosted Community Edition: you download it, run it on your own infrastructure (a VPS, Docker container, or your existing cloud), and pay n8n nothing. Unlimited workflows, unlimited executions. A few enterprise features (like SSO and advanced permissions) are reserved for paid licenses, but the automation engine itself is complete. n8n Cloud: the hosted version. n8n runs the servers, handles updates, uptime, and security patching, and you pay a monthly subscription with tiers based mainly on execution volume and features. Prices change, so check n8n's site, but entry tiers are comparable to mid-market automation tools. The unusual part, and why n8n's economics are so different from Zapier's or Make's: on self-hosted n8n, an execution costs you nothing marginal. A workflow that runs a million times costs the same as one that runs ten times: whatever your server costs.
What Self-Hosting Actually Costs
Free software still has a total cost of ownership. Here's the honest breakdown:
- Server: a modest VPS ($10-50/month) handles most small and mid-sized workloads comfortably; heavy AI-agent workloads may want more
- Setup: a Docker-comfortable person gets n8n running in an afternoon; production hardening (HTTPS, backups, authentication, monitoring) is a day or two more
- Maintenance: updates ship frequently; budget a couple of hours a month for upgrades and the occasional breaking-change fix
- The failure scenario: when the server falls over at a bad time, someone on your team has to notice and fix it. That someone is the real cost of self-hosting.
- Realistic total for a small team: $20-60/month in infrastructure plus a few hours of technical time, regardless of how many executions you run
Self-Hosted vs Cloud: How to Choose
The decision is less about money than about who carries the operational load:
- Choose self-hosted if you have technical capacity, high execution volumes, data residency requirements, or AI workflows that would be expensive on metered platforms
- Choose n8n Cloud if you want n8n's power without owning uptime: no server, no patching, support included
- Choose Cloud first, migrate later: a legitimate path is proving your workflows on Cloud, then self-hosting once volume makes the math compelling
- Stay on Zapier or Make instead if nobody on the team is technical and your volumes are modest: n8n's flexibility is wasted if building in it is a struggle
Why High-Volume and AI Workflows Change the Math
For occasional automations, platform choice barely matters financially. Two workload types break that rule. High-frequency operational workflows: lead routing, data sync, and document processing that run thousands of times a day get expensive fast on per-task or per-operation pricing. On self-hosted n8n they cost nothing extra. AI agent workflows are the sharper case. Agents loop: they call an LLM, evaluate, call a tool, call the LLM again. A single agent run might execute dozens of steps, and a metered platform bills every one. n8n has first-class AI and agent nodes, and self-hosting means those loops run free (you pay only your LLM API costs). This is why n8n has become the default recommendation for AI-heavy automation, and it's what we build most client AI workflows on.
The Bottom Line on n8n Pricing
Genuinely free exists here, which is rare in this market. The catch is not hidden fees, it's operational responsibility.
- n8n Community Edition: $0 forever, unlimited executions, you carry the ops burden
- n8n Cloud: managed convenience at a subscription, still generous on volume compared to per-task platforms
- Under ~1,000 runs a month with no technical staff: Zapier or Make's convenience probably beats n8n's economics
- 10,000+ runs a month or AI agents: self-hosted n8n is usually the cheapest serious option by a wide margin
- Either way: the platform is a fraction of the real cost. Designing the right workflows is where the value is won or lost.
Key Takeaway
So: is n8n free? Self-hosted, yes, truly, and at scale that makes it one of the best economic decisions in automation. Just price in the server and the ops time honestly, and pick Cloud if nobody wants to carry a pager for your workflows. If you're weighing n8n against what you run today, our free Ops Diagnostic can include the migration math: bring your current automation bill and your heaviest workflows, and we'll tell you whether switching is worth it.