n8n Self Hosted vs Make: Real Costs for Solo Automation
A $200 Mistake in My Automation Pipeline
Last year, an automated YouTube script pipeline I built burned through 40,000 operations on Make in less than five days. A rogue retry loop tied to an OpenAI API call kept firing every time a network request timed out. By the time I woke up in Seoul and checked my dashboard, my entire monthly operations allowance was gone, and my active workflows had stalled out.
That incident forced me to rethink how I handle automation infrastructure for my blog networks and video channels. When you run multiple content pipelines alone, automation isn’t a neat extra tool—it is your operational engine. Choosing between n8n self hosted vs make isn’t just about picking an interface you like. It comes down to how your workflows consume resources and whether your budget can withstand sudden volume spikes.
Understanding the Cost Models: Operations vs Executions
The fundamental difference between these two platforms lies in how they bill you for work. If you misunderstand this difference, you will end up paying significantly more than necessary as your operations grow.
How Make Charges: Every Step Has a Price Tag
Make (formerly Integromat) charges based on individual operations. An operation is triggered almost every time a module in your scenario performs an action. If your scenario fetches an RSS feed, splits it into five articles, sends each text block to ChatGPT, generates an image in Midjourney, synthesizes audio via ElevenLabs, and creates a draft in WordPress, that single workflow run could easily consume 15 to 20 operations.
Make offers a free tier with limited operations per month. Paid plans generally start in the $9 to $20 per month range for basic operation quotas, with costs scaling upward as your operations increase into the hundreds of thousands. Because pricing tiers change frequently, always check Make’s official pricing page for current numbers before committing your stack.
How Self-Hosted n8n Charges: Fixed Infrastructure Costs
Self-hosted n8n operates on a completely different philosophy. You download the open-source software and run it on your own Virtual Private Server (VPS), such as a basic droplet on DigitalOcean, Hetzner, or AWS Lightsail.
In n8n, billing isn’t tied to individual module steps or total operations. Instead, n8n measures full workflow executions. On a self-hosted instance, you get unlimited workflow executions and unlimited steps. Your only direct cost is the monthly server rental, which usually falls in the $5 to $20 per month range depending on the RAM and CPU specs you choose.
Keep in mind that n8n also offers a managed cloud service with tier-based pricing if you do not want to manage a server. Be sure to check n8n’s official pricing page to contrast their cloud tiers against the cost of hosting it yourself.
Setting Up Self-Hosted n8n: The Reality for Solo Founders
When people hear “unlimited automation for $5 a month,” they jump on self-hosted n8n immediately. But server management is not a free lunch. You pay for it in setup time and administrative maintenance.
Setting up n8n yourself typically involves:
- Renting a VPS with at least 2GB of RAM (1GB can work initially, but RAM spikes during large payload handling will crash your server).
- Setting up Docker and Docker Compose to run the n8n container reliably.
- Configuring a reverse proxy like Caddy or Nginx to handle SSL certificates (HTTPS) so webhooks from external services work properly.
- Setting up automated database backups (using PostgreSQL) so you do not lose your scenario configurations if the server drops.
When I set up my first n8n instance on a cheap VPS, I forgot to configure auto-pruning for execution history logs. Within two weeks, the database filled up the server’s disk space, crashing the instance overnight. I lost three days of automated blog distribution before I realized what had happened. If you are not comfortable logging into a terminal via SSH or reading Docker logs, those hours spent troubleshooting are a real operational cost.
Make Cloud: Unmatched Ease at a Variable Cost
Despite moving my high-volume workflows to n8n, I still keep a paid Make account active. Why? Because the developer experience and integration library on Make remain top-tier for rapid prototyping.
Where Make Wins for Solo Businesses
- Visual Debugging: Make’s visual execution history is clearer than n8n’s. You can inspect the exact data bundle entering and leaving every module during any past execution.
- Data Mapping: Mapping complex JSON objects between apps like Notion, Airtable, and Canva is exceptionally intuitive in Make’s drag-and-drop UI.
- Pre-built Connectors: Make has official integrations for hundreds of niche SaaS tools that might require custom HTTP requests or webhooks inside n8n.
When I want to test a new automated newsletter concept or connect a quick Google Sheet to a Slack alert, I often build it in Make first. It takes five minutes, requires zero infrastructure checks, and works instantly.
n8n Self Hosted vs Make: Side-by-Side Comparison
| Feature | Make Cloud | Self-Hosted n8n |
|---|---|---|
| Cost Structure | Metered by operations | Fixed VPS server cost |
| Setup Time | Instant (0 minutes) | 1–3 hours (requires Docker/VPS) |
| Data Volume Limits | Capped by plan tier | Limited only by server hardware |
| Maintenance | Zero (Fully managed) | Manual (Updates, backups, disk space) |
| Code Flexibility | Basic JavaScript/Functions | Native JavaScript & Python nodes |
Where Each Platform Fails
No tool is perfect, and ignoring their weaknesses will lead to broken pipelines down the road.
Make’s Weak Points
Make becomes prohibitively expensive when handling bulk data processing. If you write a scenario that loops over 500 database rows every hour, you will consume 12,000 operations a day just on data checking. Furthermore, debugging complex webhooks in Make can sometimes lead to silent scenario pauses if error handling modules are not explicitly constructed for every single step.
Self-Hosted n8n’s Weak Points
Self-hosted n8n requires you to act as your own system administrator. If n8n releases a security patch, you must pull the new Docker image and restart the container yourself. Additionally, community-made nodes in n8n can occasionally break or become unmaintained when target APIs update, forcing you to rewrite those connections using generic HTTP Request nodes instead.
My Take: Which One Should You Choose?
If you are running high-volume AI workflows—such as scraping articles, processing long-form video transcripts through Claude, generating multiple image variations, or running scheduled batch updates—self-hosted n8n is the clear winner for solo budgets. Once your infrastructure is running stably, handling 500,000 steps a month on n8n costs the exact same as handling 500 steps.
However, if your business generates healthy revenue, relies on standard integrations, and you despise system administration, Make is worth every penny. The time you save avoiding VPS configuration and database maintenance can be spent directly on content creation or marketing.
My current setup in my Seoul office is a hybrid approach: I run all lightweight, critical customer-facing triggers on Make, while all heavy content creation pipelines, feed parsers, and AI batch transformations run on a $10/month self-hosted n8n server. This keeps my monthly software overhead low while giving me unlimited ceiling for content volume.
FAQ
Is self-hosting n8n hard for someone without a coding background?
It has a learning curve. While you do not need to write complex code to build workflows in n8n, installing it requires basic comfort with terminal commands, SSH keys, Docker, and domain DNS setup. If those terms feel intimidating, consider starting with n8n’s cloud plan or Make before attempting self-hosting.
Will self-hosted n8n save money compared to Make?
If your workflows perform tens of thousands of operations per month, yes, self-hosted n8n will save you significant money. A heavy workflow schedule that might cost $100+ per month on Make can run on a $5 to $10 per month self-hosted server with n8n. However, if you only run light workflows a few times a day, Make’s free or basic tier may actually be cheaper and easier.
What server specs do I need to run self-hosted n8n reliably?
For small to medium automation workloads, a server with 2 vCPUs and 2GB to 4GB of RAM is recommended. While n8n can start on a 1GB RAM server, memory spikes during execution of large JSON files or image processing can cause the process to crash unexpected. Always configure swap space and automated database maintenance.
