n8n hosting decision
Self-hosted n8n vs n8n Cloud: decide by operational risk.
Self-hosted n8n can be the right answer, but only if you want to own the production checklist. n8n Cloud is the better comparison when webhook routing, backups, SSL, upgrades, or workers are the recurring source of failure.
Short Answer
Stay self-hosted when infrastructure control is valuable and you can verify persistence, backups, public URLs, queue workers, and upgrades. Move to n8n Cloud when n8n is still the right workflow builder but operations are taking attention away from automation work.
Production Gate Comparison
All hosting options| Gate | Self-hosted n8n means | n8n Cloud is better when | Evidence to keep |
|---|---|---|---|
| Webhook reliability | You own DNS, SSL, reverse proxy, WEBHOOK_URL, forwarded headers, and uptime. | Use cloud when public callback reliability matters more than custom network control. | External curl reaches the production URL and creates one execution. |
| Data persistence | You must preserve volumes, database, encryption key, and binary data. | Use cloud when losing workflows or credentials would be unacceptable and backup ownership is not the goal. | A restore test opens workflows and decrypts credentials. |
| Queue mode and workers | You own Redis, worker env parity, binary data access, and version alignment. | Use cloud if scaling operations are recurring distractions. | Worker processes a job, uses the same encryption key, and reads required binary data. |
| Upgrade risk | You own snapshots, rollback, migrations, release notes, and smoke tests. | Use cloud if the team wants n8n but not the upgrade playbook. | A pre-upgrade backup and rollback path have been tested. |
| Cost clarity | Server cost can look lower, but operations time is real cost. | Cloud can be cheaper when it removes repeated incidents and maintenance. | Estimate monthly server cost plus incident hours, not server cost alone. |
Stay self-hosted if
- You need private network placement, custom infrastructure control, or server-level integration.
- You can test backup restore before production data matters.
- You have someone accountable for Docker, DNS, SSL, Postgres, Redis, and upgrades.
Move managed if
- Most failures are caused by hosting operations rather than workflow logic.
- The business needs working automations more than infrastructure control.
- Webhook availability, credential stability, and upgrade safety are higher-value than server ownership.
Before You Move
Export workflows, list credentials, identify external callback URLs, capture current environment variables without secrets, and write smoke tests for the top three production workflows. A move is successful only when the same trigger, payload, output, and response timing pass in the target setup.
Generate migration brief