So you want workflow automation and AI agents without paying monthly subscription fees or managing servers? I built a complete guide for deploying n8n on Google Cloud Run - serverless, scalable, and surprisingly cheap.
Most n8n self-hosting guides throw you into Kubernetes configs or Docker Compose setups that need babysitting. This approach uses Cloud Run's serverless platform, which means:
- Costs £2-12/month (mostly database)
- No Docker builds required
- Scales to zero when idle
- Terraform option for one-command deployment
It's cheaper than n8n's managed tier and way simpler than running your own infrastructure.
Admittedly, my guide started as a custom Docker image approach, but after feedback from the actual Google Cloud Run team in, we simplified it dramatically. Now you can deploy directly using n8n's official image. I'm pretty chuffed to have PRs from the actual GCR team in SF.
I've run this reliably for months and the guide has over 400 stars on GitHub now.
(Tip: if you want it really cheap, then you can swap out Google's Cloud SQL for Neon.tech's free tier).
IMO this setup is perfect for:
- Personal (AI) automation workflows
- Small team deployments
- Learning Cloud Run and serverless platforms
- Keeping costs predictable and low
Less ideal for:
- High-availability multi-region deployments
- Workflows regularly exceeding 60-minutes execution time
- Organisations needing dedicated support
My repo has everything you need. If you have an questions or run into any issues, the repo has a discussions section, and I'm happy to help troubleshoot. PRs always welcome!
Top comments (0)