🟢 All systems operational · Self-hosted automation on hardware you rent
+ Add Provider
Workflow Automation · Self-hosted

Self-hosted n8n —
Your Workflows, Your Box

n8n is a workflow automation tool you can run yourself. Self-hosting moves the credentials, the data travelling through each workflow and the whole execution history onto a server you rent — and turns the bill into a monthly server price instead of a per-execution one.

See what it costs to host → Browse all VPS plans
n8n — docker
$ docker run -d --name n8n -p 5678:5678 \
    -v n8n_data:/home/node/.n8n docker.n8n.io/n8nio/n8n
✓ editor on http://localhost:5678
# production: swap SQLite for Postgres
$ export DB_TYPE=postgresdb
$ export EXECUTIONS_MODE=queue
✓ workers taking jobs from redis

Why run it yourself

A visual editor, not a DSL
Workflows are built by wiring nodes together, and every run can be opened step by step to see exactly what each node received and returned.
Credentials never leave your server
API keys, tokens and the payloads moving through a workflow sit in your own database. n8n encrypts credentials with a key you set and must keep — lose it and they cannot be read back.
AI nodes in the same canvas
n8n ships nodes for calling language models and building agents, so a workflow can talk to a model you host yourself instead of a metered API.
Fair-code, and n8n says so
The Sustainable Use License lets anyone self-host and modify n8n for their own internal business or personal use, and restricts reselling it as a service. n8n states plainly that this is not an OSI open-source licence.
Webhooks and schedules
A workflow starts from an incoming webhook, a timer, or by hand. A webhook trigger is why n8n usually wants a public address and a certificate.
Scales by adding workers
Queue mode splits the editor from the machines that execute workflows: Postgres holds the data, Redis hands out the jobs, and throughput grows by starting more worker containers.

What actually runs on the box

These four steps are ours, not n8n’s. n8n publishes no per-workflow hardware requirement, and it could not: the load comes from what your workflows do — a nightly sync of 200 rows and a webhook taking 50 requests a second are the same product and completely different machines. So instead of inventing a minimum, we price the ladder and say what sits on each rung. What is not ours: n8n stores data in SQLite by default, wants PostgreSQL for production, and requires it for queue mode — that is n8n’s own documentation.
Setup What is running Our sizing step
Trying it out One n8n container, SQLite on a mounted volume 1 core / 2 GB
Small production n8n plus PostgreSQL, a fixed encryption key, persistent storage, a certificate for the webhook address 2 cores / 4 GB
Queue mode n8n main process, PostgreSQL, Redis as the job broker, one or two worker containers 4 cores / 8 GB
Busy queue mode The same, with more workers and room for large payloads and execution history 8 cores / 16 GB

Two things this table deliberately does not say. It does not promise a number of executions per second — that depends on the nodes in your workflow, not on n8n. And it does not cover the database once you move Postgres to its own machine, which is the usual next step after queue mode.

Cheapest server at each step

All VPS plans →

For every rung of the ladder above, the cheapest VPS in the DohoHub catalogue that meets it right now. Prices are the ones our feed already tracks, normalised to compare and shown in the currency picked in the header. One machine often wins two rungs at once — when it does, it appears once and the note says which rungs it covers.

Provider Plan Specs Price Visit
Database Mart
Express Linux VPS
cheapest from 1 core / 2 GB RAM up to 2 cores / 4 GB RAM
 2 vCPU4 GB RAM60 GB SSD
$2.88/mo
 Visit
Contabo
Cloud VPS 12 (12 GB RAM)
cheapest at 4 cores / 8 GB RAM
 6 vCPU12 GB RAM100 GB NVME0.2 Gbps
$3.93/mo
renews at $5.24
 Visit
Contabo
Cloud VPS 24
cheapest at 8 cores / 16 GB RAM
 8 vCPU24 GB RAM200 GB NVME0.3 Gbps
$6.11/mo
renews at $8.14
 Visit

These are ordinary VPS plans, not an n8n product — no provider here sells “n8n hosting”, and installing it is on you. Where a provider’s headline price only buys an introductory term, the renewal rate is printed beneath it. Check the total on the provider’s own checkout before you buy.

Pick the box, then install it once

Filter the whole catalogue by cores, memory, storage and location — then run the Docker command above.

Find a server by specs →
↑