Hosting & Pricing Playbook
Every workflow in this guide runs on n8n Cloud by default, but the moment you're running client workloads or hitting execution limits, hosting choice becomes a real cost and reliability decision — not just a settings toggle.
Cloud vs. self-hosted, at a glance
| n8n Cloud | Self-hosted | |
|---|---|---|
| Setup time | ~3 minutes | 3 minutes to several hours, depending on comfort with servers |
| Cost | €20–€667+/month by tier | Community edition is free; you pay only for a small VPS (from a few dollars a month) |
| Uptime | Managed for you | Your responsibility — if your server or home connection drops, workflows stop |
| Maintenance | None | You apply updates and manage backups yourself |
| Execution limits | Capped by plan tier | None imposed by n8n — bounded only by your server's resources |
Running mission-critical client automation from a personal laptop is not a real hosting strategy — if the laptop closes, the internet drops, or the machine restarts, every workflow depending on it stops silently. Either n8n Cloud or a small always-on VPS solves this; a home computer generally does not.
Current n8n Cloud pricing
| Plan | Price (annual billing) | Executions/mo | Notes |
|---|---|---|---|
| Starter | €20/mo | 2,500 | 5 concurrent executions, unlimited users, 2,300 AI credits/mo |
| Pro | €50/mo | 10,000 | Admin roles, global variables, workflow history, 13,700 AI credits/mo |
| Business | €667/mo | 40,000 | SSO/SAML, Git version control, AI Assistant, self-hosted deployment option |
| Enterprise | Custom | Custom | SLA support, log streaming, extended data retention |
A Startup plan offers roughly 50% off the Business tier for teams under 20 people. A free, self-hosted Community Edition is available directly from n8n's open-source repository at no license cost — you provide the server.
What self-hosting actually takes
Self-hosting bypasses execution caps entirely and can bring your effective cost under $5/month, but it shifts uptime and maintenance onto you. Every Google integration used in this guide also needs its own OAuth setup through the Google Cloud Console when self-hosted, rather than the one-click "sign in with Google" n8n Cloud provides — budget real setup time for this the first time through, even though it only needs doing once per Google API.
Self-hosting on a VPS
Any VPS provider that offers a one-click n8n install (Hostinger and several others do) gets you running in minutes rather than hours: pick a server location close to your primary users to minimize latency, choose the entry-tier plan while testing, and generate a strong root password through the provider's own tooling rather than reusing one. Budget a recurring reminder to check for and apply n8n version updates — self-hosted instances don't update themselves.
The margin math for agencies
If you're running automation for multiple clients, self-hosting is usually the better economics once you pass roughly 2–3 active client workflows: a $5–15/month VPS handling every client's execution volume with no per-execution ceiling beats paying for Cloud tier upgrades every time a new client pushes you over a plan's cap. Pass the infrastructure cost through in your retainer pricing (see Layer 10) rather than absorbing it — it's a real, recurring cost of delivery.
Choosing for yourself vs. a client
For your own learning and early portfolio builds: n8n Cloud's free trial or Starter tier removes every infrastructure question so you can focus entirely on the concepts in Layers 00–09. For paying client work: self-host once you have two or more active clients, or immediately if a client's compliance requirements demand it — some industries (finance, healthcare) require self-hosting by default regardless of cost.
Sources
| Source | Used for |
|---|---|
| n8n.io official pricing page | Current Cloud plan pricing and execution limits |
| NovelVista, "n8n Pricing 2026: Complete Guide" | Self-hosted cost comparison context |
| ExpressTech, "The Real Cost of Self-Hosting n8n in 2026" | VPS hosting cost baseline |
Pricing changes — verify current figures at n8n.io/pricing before quoting a client.