Self-hosting n8n: what you need and why run it on your own server

By the serverserv team · 5 min read · Updated 15 October 2026

n8n is a popular tool for automating work between apps, and you can run it on your own server instead of a hosted service. This guide explains why people self-host n8n, what a reliable setup needs, and how to choose a server size.

What n8n is

n8n is a visual workflow automation tool. You connect apps and services with drag-and-drop nodes, and each workflow runs when something triggers it: a schedule, a webhook, a new form entry or a new email, for example.

People use it to send new leads from a form to a CRM and an email list, pull data into scheduled reports, post alerts to a chat channel, and process documents. Its AI nodes let you build chatbots, agents and document flows with the AI model provider of your choice.

Why run n8n on your own server

• No per-execution pricing from your host. Hosted automation services often price by tasks or executions. On your own server, what you can run depends on your CPU and RAM. serverserv does not count or cap workflows or executions.
• Your data stays on your server. Workflows, execution history and saved credentials are stored on the server you control.
• Room for heavy work. Large data transfers, long-running workflows and AI steps get the full resources of your plan.
• Freedom to connect anything. Call internal APIs, databases and webhooks from a server you manage.

Self-hosting also lets you choose when to update. You can test a new n8n version before you rely on it, instead of having changes arrive on someone else's schedule. When the time comes, follow how to update self-hosted n8n without losing workflows.

The trade-off is that a self-hosted tool needs looking after. The next section covers what that involves.

What a self-hosted n8n needs

• A domain and HTTPS. Webhooks and many app integrations need a public address that other services can reach, and most require HTTPS. Point a subdomain, such as automate.yourcompany.com, at your server. You can register a domain on the domains page.
• A database. n8n stores workflows, credentials and execution history in a database. SQLite is the default; PostgreSQL is a common choice for heavier production use.
• The encryption key. n8n encrypts saved credentials with a key. Keep a safe copy of it: if you move or restore n8n without it, your saved credentials cannot be decrypted.
• Backups. Back up the database and the encryption key regularly. Exporting important workflows as JSON is an easy extra safety net.
• Updates. n8n releases new versions often. Read the release notes and take a backup before you update.
• Housekeeping. Execution history grows over time. Set n8n to prune old executions so the database and disk do not fill up.
• Strong access control. Use a strong password for your n8n account and limit who has access.

A note on n8n's licence

n8n is distributed under a fair-code licence. Its source code is available and many people self-host it, but it is not a standard open source licence, and it places conditions on some commercial uses. If you plan to use n8n commercially, especially to offer it to others as part of a product or service, check n8n's own licence terms first.

Choosing an n8n plan

Every serverserv n8n plan is a dedicated KVM VPS with NVMe storage, n8n pre-installed, setup included and unlimited workflows and executions. Prices are per month.

• n8n Personal: 4 vCPU, 8 GB RAM, 100 GB NVMe. For one person or a small business with a few active workflows. From $17.99/mo on a 3 year term, or $24.99 monthly.
• n8n Pro: 6 vCPU, 12 GB RAM, 200 GB NVMe. For heavier use and AI workflows. From $28.49/mo on a 3 year term, or $39.99 monthly.
• n8n Team: 12 vCPU, 48 GB RAM, 400 GB NVMe. For teams with many workflows or high volume. From $57.49/mo on a 3 year term, or $79.99 monthly.
• n8n Enterprise: 18 vCPU, 96 GB RAM, 600 GB NVMe. For the largest teams and volumes. From $114.99/mo on a 3 year term, or $159.99 monthly.

Workflows that handle large files, many items at once or long AI steps use more memory, so lean towards a bigger plan if that describes your work. A handful of simple workflows that run every few minutes need far less than a busy webhook that receives thousands of requests a day. Start with the plan that fits your current workflows, watch how the server copes, and request an upgrade from your client area when you need more room.

Getting started

1. Order an n8n plan on the n8n hosting page.
2. When your server is ready, your login details are emailed to you and shown in your client area. Open n8n in your browser and start building. You do not need to know Linux.
3. Moving from another n8n? Export your workflows as JSON and import them on your new server. Open a ticket if you want help.
4. For AI workflows, connect your own AI provider account, for example with an API key. AI model costs are not included in the plan.

If you are unsure how updates or backups work on your plan, ask our team before you rely on the server for important work.

Questions