Run the n8n workflow automation platform on a private server you control. Instead of paying per execution, you pay a flat monthly price for the VPS and run as many workflows as it can handle. Ubuntu with Docker, NVMe storage, DDoS protection and 25+ locations to keep n8n close to your APIs and your data.
♾️ No execution limits🐳 Docker-ready💾 NVMe SSD🌍 25+ locations💳 No KYC
Germany and Netherlands 2.0 plans include 20 TB of traffic per month; other plans are unmetered.
How pricing works
Prices are per month for the billing term selected on the card (3 mo by default). Switch the term and the price and total update. The longer the term, the lower the monthly rate: 3 mo — 5%, 6 mo — 10%, 12 mo — 15%. The order button opens checkout with the selected term.
n8n VPS requirements
Our sizing guidance for self-hosted n8n. Memory use grows with parallel executions, large binary files and AI agent workflows.
Setup
RAM
vCore
Disk
Plan to start with
Testing and personal automations (SQLite)
2 GB
1
25 GB
Lite 2.0 — $10/mo
Team production with PostgreSQL and webhooks
4 GB
2–3
40 GB
Lite++ 2.0 — $17/mo
Queue mode: main + workers + Redis
8 GB
4–6
60 GB
Pro+ 2.0 — $27/mo
Heavy AI agents, many parallel executions
12–16 GB
8–12
80–100 GB
Pro++ / Super+ 2.0
Prices shown for Germany (1-month term). Running local LLMs next to n8n needs a GPU server.
How to self-host n8n on a VPS
Order a VPS with Ubuntu
Pick a location near the services you automate and choose Ubuntu 24.04 LTS at checkout. You get the IP and root password in about 5 minutes.
Point a domain to the server
Create an A record such as n8n.example.com pointing to the VPS IP. n8n needs HTTPS for OAuth sign-ins and most webhooks.
Install Docker and start n8n
Install Docker, then run n8n with Docker Compose together with PostgreSQL and a reverse proxy such as Caddy or Traefik that issues a free Let's Encrypt certificate.
Secure and back it up
Set N8N_HOST and WEBHOOK_URL, enable a firewall, keep n8n updated and schedule database backups or order server snapshots.
n8n Cloud vs self-hosted n8n on a VPS
n8n Cloud
Self-hosted n8n on LikeVPS
Pricing
Subscription tiers with execution limits
Flat VPS price from $10/month
Executions and workflows
Capped by the plan
Limited only by server resources
Data location
Chosen by n8n
Your choice of 25+ countries
Custom code and community nodes
Restricted to what the cloud allows
Full control: npm packages, custom nodes, env variables
Maintenance
Managed for you
You update n8n (one Docker command)
The n8n Community Edition is free to self-host under n8n’s Sustainable Use License; some enterprise features require a paid n8n license.
Why host n8n with LikeVPS
♾️
No per-execution billing
Run thousands of workflow executions a day for the same monthly price.
🔐
Your data stays yours
Credentials, payloads and logs live on your server in the country you choose — useful for GDPR and client data.
🐳
Ready for Docker
KVM virtualization with full root access: Docker, Docker Compose, Redis and PostgreSQL work out of the box.
🌍
Close to your APIs
25+ locations, so webhooks and API calls travel the shortest route to the services you automate.
🛡️
DDoS protection
Public webhook endpoints are protected by network-level filtering at no extra cost.
📈
Grow without migrating
Upgrade RAM and CPU in minutes when you switch to queue mode or add AI agents — no reinstall.
Frequently asked questions
How much RAM does n8n need?
n8n itself runs in 1–2 GB of RAM for light use. For a team instance with PostgreSQL and steady webhook traffic we recommend 4 GB; queue mode with separate workers and Redis works best with 8 GB or more.
Is self-hosted n8n free?
The n8n Community Edition can be self-hosted for free under n8n’s Sustainable Use License. You only pay for the server. Some enterprise features, such as SSO and advanced permissions, require a paid license from n8n.
Can I install n8n with Docker on your VPS?
Yes. All plans use KVM virtualization with full root access, so Docker and Docker Compose work without restrictions. The official n8n Docker image is the recommended way to install and update it.
Which location should I choose for n8n?
Place n8n close to the APIs and databases it talks to most. For European companies, Germany or the Netherlands also keeps data inside the EU; for US-based SaaS tools choose the USA.
Can n8n run AI agents on this VPS?
Yes — AI agent workflows that call OpenAI, Anthropic, Google or other model APIs run fine on a 4–8 GB plan. Running open-source models locally (for example with Ollama) needs much more memory or a GPU; see our GPU servers.
How do I update self-hosted n8n?
Pull the new Docker image and restart the container — usually a one-minute job. Back up the database first and read the n8n release notes for breaking changes.
Do you back up my n8n server?
Backups are your responsibility by default. You can schedule PostgreSQL dumps to external storage and order server snapshots as an add-on.
n8n hosting: why run n8n on your own VPS
n8n is an open, node-based workflow automation tool — a self-hostable alternative to Zapier and Make. You can use n8n’s own cloud, or you can self-host n8n on a VPS and keep full control over data, cost and integrations.
Benefits of self-hosting n8n
Predictable cost — a fixed monthly server price instead of plans built around executions.
Data control — API keys, customer data and logs stay on your server in a jurisdiction you pick.
Flexibility — install community nodes, npm packages and custom code nodes, set any environment variable, connect to databases inside your private network.
Performance — scale vertically in minutes and enable queue mode with workers when traffic grows.
Recommended n8n stack on a VPS
Ubuntu 24.04 LTS, Docker Compose, the official n8n image, PostgreSQL as the database and Caddy or Traefik as a reverse proxy with automatic HTTPS. For high volumes add Redis and run n8n in queue mode with one or more worker containers.
What else fits on the same server
A 4–8 GB server usually has room for related tools: a small PostgreSQL database for your workflows, a vector database such as Qdrant for AI agents, Uptime Kuma for monitoring or a Baserow/NocoDB instance as a lightweight back office.
