Key takeaways (updated April 17, 2026):
- n8n Cloud starts at $24/mo for 2,500 executions (Starter). Pro is $60/mo for 10,000. Enterprise is custom.
- Self-hosted n8n Community Edition is free with unlimited executions - you just need a server ($3-7/mo).
- n8n counts executions per workflow run, not per step - much cheaper than Zapier and Make at scale.
- For most solo developers and small teams, self-hosting saves $200-700/year vs Cloud.
- n8n AI agent workflows run on the same execution model - no separate AI pricing tier.
n8n is one of the best workflow automation tools available right now. 400+ integrations, a visual editor, full code access when you need it, and an active open-source community. But the n8n cloud pricing in 2026 can be confusing - especially when you factor in self-hosted options.
What Changed in n8n Pricing (2026)
n8n made significant pricing changes in late 2025 that are still causing confusion in 2026. Here’s what happened:
- Self-hosted Community Edition stayed free - unlimited executions, no license fees. This hasn’t changed.
- Self-hosted Enterprise got a price increase - n8n introduced tiered pricing for self-hosted Enterprise features (SSO, LDAP, audit logs). This sparked a heated community discussion with users concerned about the direction.
- Cloud plans were restructured - the old Free tier was removed. Cloud now starts at $24/mo (Starter) with 2,500 executions. The previous free tier with 5 active workflows is gone.
- Execution counting changed - n8n shifted to counting “successful executions” only. Failed and test runs don’t count toward your limit. This is actually more generous than before.
Bottom line: If you’re using the free Community Edition for self-hosting, nothing changed for you. If you were on the old free Cloud tier, you now need to pay $24/mo or switch to self-hosting. If you need Enterprise self-hosted features, the cost went up.
There’s n8n Cloud (the hosted version starting at $24/mo), n8n Community (the free self-hosted version), and then a whole range of third-party hosting options at wildly different price points. Some charge per execution. Some charge flat. Some are free but cost you hours of setup time.
I’ve used n8n across multiple setups over the past year. Here’s every n8n pricing option explained - what you actually pay, what you get, and which one makes sense for your use case.
How n8n Counts Executions (Important)
Before diving into pricing, understand how n8n counts executions - it’s the key difference between n8n and competitors like Zapier and Make.
One execution = one workflow run, regardless of how many steps are in that workflow. A 3-step workflow and a 30-step workflow both count as one execution.
Compare that to Zapier, where each step in a Zap counts as a separate “task.” A Zap with 5 steps uses 5 tasks per run. The same automation on n8n uses 1 execution.
This means n8n’s per-execution cost is significantly lower than it looks on paper:
| Platform | 5-step workflow, 1,000 runs/mo | Cost |
|---|---|---|
| n8n Cloud Starter | 1,000 executions | $24/mo |
| Zapier Professional | 5,000 tasks | $49/mo |
| Make Core | 10,000 operations | $10.59/mo |
Make counts per “operation” (each module = 1 operation), so a 10-step scenario uses 10 operations per run. Make looks cheap on paper but can burn through operations fast with complex scenarios.
The takeaway: When comparing n8n pricing to Zapier or Make, multiply your Zapier task count or Make operation count by the number of steps to get the equivalent n8n execution count. n8n almost always wins on per-unit cost for multi-step workflows.
n8n Cloud Pricing 2026
n8n Cloud is the easiest way to get started. n8n hosts everything for you - no server, no Docker, no maintenance. You sign up, build workflows, and they handle the infrastructure.
Here’s what each plan costs in 2026:
| Plan | Price (monthly) | Price (annual) | Executions/mo | Active Workflows | Users |
|---|---|---|---|---|---|
| Starter | $24/mo | $20/mo | 2,500 | 5 | 1 |
| Pro | $60/mo | $50/mo | 10,000 | 50 | 3 |
| Enterprise | Custom | Custom | Unlimited | Unlimited | Unlimited |
Annual billing saves about 17% across all plans. If you’re committing to n8n Cloud, the annual price is worth it.
What’s included on all plans:
- Managed hosting (no server to maintain)
- Automatic updates
- Built-in credentials encryption
- Community nodes support
What to watch out for:
The execution limits are the real constraint. An “execution” is one workflow run - so if you have a workflow that triggers every 5 minutes, that’s 8,640 executions per month from a single workflow. The Starter plan’s 2,500 limit would be blown in 9 days.
The Pro plan’s 10,000 executions works for light automation, but if you’re running webhook-triggered workflows with any real traffic, you’ll hit the ceiling fast.
n8n Cloud pricing per execution:
- Starter: $0.0096/execution ($24 / 2,500)
- Pro: $0.006/execution ($60 / 10,000)
- Overage: additional execution packs available, pricing varies
For comparison, Zapier charges $0.01-$0.03 per task depending on your plan. n8n Cloud is cheaper per execution, but the base price is higher than Zapier’s Starter ($19.99/mo for 750 tasks).
n8n Community Edition (Self-Hosted - Free)
The community edition is the same n8n software, minus some enterprise features (SSO, LDAP, advanced permissions). For most solo developers and small teams, the community edition has everything you need.
What’s free:
- Unlimited workflows
- Unlimited executions
- All 400+ integrations
- Full API access
- No user limits
What you still need:
- A server to run it on
- Basic knowledge of Docker or npm
- Time for maintenance (updates, backups, SSL)
The community edition is genuinely free - no freemium tricks, no execution limits, no “upgrade to unlock this node.” The catch is hosting. You need a server, and that server costs money (or time, which is also money).
Self-Hosted n8n: What It Actually Costs
This is where most people get confused. “Free” software still needs to run somewhere. Here’s what every hosting option actually costs:
1. InstaPods - $3/mo (managed)
One-click deploy. n8n comes pre-installed on a real Linux server with HTTPS, custom domains, and SSH access. No Docker setup, no nginx config, no SSL certificates to manage.
- Cost: $3/mo (Launch plan) or $7/mo (Build plan - recommended for heavier workflows)
- Setup time: About 60 seconds
- You manage: Nothing. Updates, backups, SSL handled automatically.
- Executions: Unlimited
The $3/mo Launch plan (0.5 vCPU, 512MB RAM) handles light workflows fine. If you’re running 10+ active workflows or doing heavy data processing, the $7/mo Build plan (1 vCPU, 1GB RAM) gives more headroom.
Deploy n8n on InstaPods - takes about 60 seconds.
2. PikaPods - ~$3.80/mo (managed)
Similar one-click approach. PikaPods runs n8n in a managed setup with automatic updates. Pricing is resource-based - you pay for the CPU and RAM you allocate.
- Cost: ~$3.80/mo (minimum resources)
- Setup time: A few minutes
- You manage: Nothing
- Executions: Unlimited
Solid option. Slightly more expensive than InstaPods, but same general concept - managed hosting without the server management overhead.
3. Hetzner VPS + Docker - $4-5/mo (self-managed)
The cheapest option if you’re comfortable with Linux. Rent a VPS from Hetzner (CX22 at $4.51/mo), install Docker, and run n8n with Docker Compose.
- Cost: $4-5/mo
- Setup time: 30-60 minutes
- You manage: Everything - OS updates, Docker updates, n8n updates, SSL, backups, firewall
- Executions: Unlimited
You get a full server you control completely. The trade-off is maintenance. You’re the sysadmin. When n8n releases an update, you pull the new image. When the SSL cert needs renewing, you handle it. When something breaks at 2am, that’s your problem.
4. Coolify on a VPS - $5-8/mo (semi-managed)
Coolify is a self-hosted PaaS that gives you a Heroku-like experience on your own VPS. Install Coolify on a $5-8/mo VPS, then deploy n8n through its dashboard.
- Cost: $5-8/mo (VPS cost, Coolify itself is free)
- Setup time: 30-60 minutes for Coolify, then a few minutes for n8n
- You manage: VPS maintenance, Coolify updates
- Executions: Unlimited
Good middle ground if you want a GUI for managing multiple apps. Coolify handles SSL, deployments, and basic monitoring. You still manage the underlying server.
5. Elestio - ~$17/mo (fully managed)
Elestio is a managed open-source hosting platform. They handle updates, backups, and monitoring. More polished than a raw VPS, but significantly more expensive.
- Cost: ~$17/mo (cheapest tier)
- Setup time: A few minutes
- You manage: Nothing
- Executions: Unlimited
At $17/mo, you’re paying close to n8n Cloud Starter pricing ($24/mo) but without the execution limits. Makes sense if you want hands-off hosting and don’t mind paying for it.
Cloud vs Self-Hosted: Side-by-Side
| n8n Cloud (Starter) | Self-Hosted (InstaPods) | Self-Hosted (Hetzner VPS) | |
|---|---|---|---|
| Monthly cost | $24/mo | $3-7/mo | $4-5/mo |
| Executions | 2,500/mo | Unlimited | Unlimited |
| Active workflows | 5 | Unlimited | Unlimited |
| Setup time | 2 minutes | 1 minute | 30-60 minutes |
| Maintenance | None | None | All on you |
| Updates | Automatic | Automatic | Manual |
| SSH access | No | Yes | Yes |
| Custom domain | Yes (paid plans) | Yes (all plans) | Yes (manual setup) |
| Data location | n8n’s servers | Your server (EU) | Your server (your choice) |
| Best for | Teams, zero-ops | Solo devs, cost-conscious | Full control |
The core trade-off: n8n Cloud saves you all operations work but limits your executions and costs 3-8x more. Self-hosted gives you unlimited everything but requires some level of server awareness.
Hidden Costs Most Guides Don’t Mention
Every n8n pricing comparison shows the headline numbers. Here’s what they leave out:
n8n Cloud hidden costs:
- Overage fees - exceed your execution limit and you’ll need to buy additional execution packs. Pricing varies but expect $10-20 for each additional 2,500 executions.
- Annual lock-in - the 17% discount requires annual billing. Monthly billing is $24/mo (Starter) vs $20/mo (annual). Canceling mid-year doesn’t get you a refund.
- Credential limits - Cloud plans limit the number of stored credentials. If you connect many services, you may need a higher plan.
- No data export on Free - there was a free tier; it’s gone. If you built workflows on it, you now pay or migrate.
Self-hosted hidden costs:
- Time - setting up a VPS with Docker, nginx, SSL, and backups takes 1-4 hours. Ongoing maintenance adds 1-2 hours/month. Value your time at $50/hr and that’s $50-200 for setup plus $50-100/month.
- Database costs - n8n defaults to SQLite, which works for light use. Production workloads need PostgreSQL, which needs its own resources or a managed database ($5-15/mo).
- Backup storage - if your workflows process important data, you need automated backups. Cloud storage for backups adds $1-5/mo.
- SSL certificates - free with Let’s Encrypt, but you need to set up auto-renewal. Managed hosting platforms handle this for you.
- Monitoring - how do you know when n8n crashes at 3am? You need uptime monitoring ($0-10/mo depending on tool).
The actual cost comparison with hidden costs included:
| n8n Cloud Starter | Self-hosted (managed) | Self-hosted (raw VPS) | |
|---|---|---|---|
| Base cost | $24/mo | $3-7/mo | $4-5/mo |
| Overage (if exceeded) | $10-20/mo | $0 | $0 |
| Time cost (setup) | $0 | $0 | $50-200 one-time |
| Time cost (monthly) | $0 | $0 | $50-100/mo |
| Database | Included | Included | $0-15/mo |
| Monitoring | Included | Included | $0-10/mo |
| Realistic total | $24-44/mo | $3-7/mo | $54-330/mo |
The counterintuitive finding: a raw VPS is the “cheapest” option on paper but the most expensive when you factor in your time. Managed self-hosting ($3-7/mo) is the cheapest option in total cost because it eliminates the time burden while keeping unlimited executions.
Which Plan Should You Pick?
Choose n8n Cloud Starter ($24/mo) if:
- You don’t want to think about servers at all
- You run fewer than 2,500 workflow executions per month
- You need team collaboration features
- Your company pays for your tools
Choose n8n Cloud Pro ($60/mo) if:
- You have multiple team members building workflows
- You need up to 10,000 executions per month
- You want priority support
- You’re running n8n for a business with budget
Choose self-hosted on InstaPods ($3-7/mo) if:
- You want unlimited executions at the lowest price
- You don’t want to manage servers yourself
- You’re a solo developer or small team
- You want SSH access for debugging or custom setups
Choose self-hosted on a VPS ($4-5/mo) if:
- You’re comfortable with Docker and Linux administration
- You want full control over the server
- You’re already managing other self-hosted apps
- You enjoy infrastructure work (or at least tolerate it)
Choose Elestio (~$17/mo) if:
- You want fully managed hosting but need unlimited executions
- The n8n Cloud price is too high but you don’t want any ops burden
- You need managed backups and monitoring included
The Execution Math
Let’s make the execution pricing concrete. Say you have these workflows:
- Slack notification bot - triggers on new messages in a channel, 200 executions/day
- CRM sync - runs every hour, 24 executions/day
- Invoice processor - webhook-triggered, 10 executions/day
- Weekly report - runs once a week, ~4 executions/month
That’s 234 executions per day, or roughly 7,020 per month.
On n8n Cloud Starter ($24/mo with 2,500 executions), you’d blow through your limit in 11 days. You’d need the Pro plan at $60/mo.
On any self-hosted option, the same workload runs without limits for $3-7/mo. That’s a $53-57/mo difference - $636-684/year.
For heavy automation users, self-hosting pays for itself almost immediately.
n8n vs Zapier vs Make: Pricing Compared
The three main workflow automation platforms use different pricing models. Here’s how they stack up for a real workload:
Test scenario: 20 workflows, averaging 5 steps each, running 500 times per month total.
| n8n Cloud (Starter) | n8n Self-Hosted | Zapier Professional | Make Core | |
|---|---|---|---|---|
| Monthly cost | $24/mo | $3-7/mo | $49/mo | $10.59/mo |
| Unit counting | 500 executions | Unlimited | 2,500 tasks | 5,000 ops |
| % of limit used | 20% | N/A | 33% (of 750) | 50% (of 10K) |
| Cost at 2x scale | $24/mo | $3-7/mo | $49/mo | $10.59/mo |
| Cost at 10x scale | $60/mo (upgrade) | $3-7/mo | $69/mo (upgrade) | $18.82/mo (upgrade) |
| Code access | Full (JavaScript/Python) | Full | Limited (Code by Zapier) | Limited (JS/Python modules) |
| Self-hosted option | No | Yes | No | No |
n8n’s advantage grows with complexity. Simple 2-step automations are comparable across platforms. But once your workflows have 5-10 steps with branching logic, n8n’s per-workflow counting and full code access pull ahead.
Make is the cheapest cloud option for simple workflows but gets expensive fast with complex scenarios (each module = 1 operation).
Zapier has the largest integration library (6,000+ apps vs n8n’s 400+) but is the most expensive per execution for multi-step workflows.
n8n AI Agent Workflows
n8n has leaned heavily into AI agent capabilities. You can build AI agent workflows using nodes for OpenAI, Anthropic, Google AI, and local models - including tool-calling agents that execute multi-step reasoning chains.
Pricing for AI workflows: There’s no separate AI pricing tier in n8n. AI agent workflows count as regular executions - one workflow run = one execution, whether it’s a simple Slack notification or a complex AI agent chain. The AI-specific costs come from the LLM API calls your workflows make (OpenAI, Anthropic, etc.), which you pay directly to those providers.
This is a significant advantage over platforms like Relevance AI or Langflow that charge per AI agent execution on top of LLM costs.
What this means for cost planning:
- n8n execution cost: same as any workflow ($0.006-0.0096/execution on Cloud, free on self-hosted)
- LLM API cost: varies (GPT-4o ~$5/1M input tokens, Claude Sonnet ~$3/1M input tokens)
- Total cost: n8n hosting + LLM API fees. No AI surcharge.
If you’re building AI agents, self-hosting n8n removes the execution cost entirely. You pay only for the LLM API calls, plus $3-7/mo for the server.
FAQ
Is n8n free?
Yes - the n8n Community Edition is 100% free and open source. You can download it and run it on your own server with no license fees, no execution limits, and no feature gates on core functionality. You just need a server to run it on, which costs $3-5/mo from most hosting providers.
How much does n8n Cloud cost?
n8n Cloud starts at $24/mo for the Starter plan (2,500 executions, 5 active workflows, 1 user). The Pro plan is $60/mo (10,000 executions, 50 workflows, 3 users). Enterprise pricing is custom.
What is the n8n pricing per execution?
On n8n Cloud Starter, each execution costs about $0.0096 ($24 / 2,500). On Pro, it’s about $0.006 ($60 / 10,000). Self-hosted n8n has no per-execution cost - you pay only for the server, regardless of how many workflows you run.
Is n8n cheaper than Zapier?
It depends on usage. n8n Cloud Starter ($24/mo for 2,500 executions) is more expensive than Zapier Starter ($19.99/mo for 750 tasks) per dollar, but cheaper per execution. Self-hosted n8n ($3-7/mo for unlimited executions) is dramatically cheaper than both for any meaningful automation workload.
How much does n8n software cost?
n8n’s cost depends entirely on how you run it. The open-source Community Edition software is free. n8n Cloud (hosted by n8n) starts at $24/mo. Self-hosting on a managed platform costs $3-7/mo for the server. Self-hosting on a raw VPS costs $4-5/mo plus your time for setup and maintenance.
Can I use n8n for free lifetime?
Yes. The n8n Community Edition is MIT-licensed open source software that’s free forever with no execution limits, no feature gates, and no time restrictions. You need a server to run it ($3-7/mo), but the n8n software itself is permanently free. There is no free tier for n8n Cloud anymore - the hosted version requires a paid plan starting at $24/mo.
Can I start free and upgrade later?
Yes. Start with the community edition on a $3/mo server, and if you outgrow it or need enterprise features (SSO, LDAP, audit logs), you can migrate to n8n Cloud or n8n Enterprise at any time. Your workflow data exports cleanly.
What server specs does n8n need?
Minimum: 1 vCPU and 512MB RAM for light workflows (under 10 active). Recommended: 1 vCPU and 1GB RAM for moderate use. Heavy use (50+ active workflows, high-frequency triggers) benefits from 2 vCPU and 2GB+ RAM.
How does n8n count executions vs Zapier tasks?
n8n counts per workflow run - one run = one execution regardless of step count. Zapier counts per step - a 5-step Zap uses 5 tasks per run. For a 5-step workflow running 1,000 times/month, n8n uses 1,000 executions while Zapier uses 5,000 tasks. This makes n8n significantly cheaper for complex multi-step automations.
Is n8n cheaper than Make?
For simple workflows, Make’s Core plan ($10.59/mo for 10,000 operations) is cheaper than n8n Cloud. But Make counts per module (step), so complex scenarios burn through operations fast. A 10-step scenario running 1,000 times uses 10,000 operations on Make but only 1,000 executions on n8n. Self-hosted n8n ($3-7/mo, unlimited) is cheaper than both in all scenarios.
How much does n8n cost in 2026?
It depends on how you host it. n8n Cloud starts at $24/mo (Starter) or $60/mo (Pro). Self-hosted n8n Community Edition is free - you just need a server, which costs $3-7/mo on managed platforms like InstaPods or $4-5/mo on a raw VPS. For most solo developers, self-hosting at $3-7/mo with unlimited executions is the best value. Enterprise pricing is custom and starts much higher.
What does n8n AI agent pricing look like in 2026?
There’s no separate AI pricing tier in n8n. AI agent workflows - including tool-calling agents with OpenAI, Anthropic, or local models - count as regular executions. One AI agent workflow run = one execution, same as a simple Slack notification. You pay your n8n hosting cost plus whatever LLM API fees you incur (GPT-4o ~$5/1M input tokens, Claude Sonnet ~$3/1M input tokens). Self-hosting removes the n8n cost entirely - you pay only LLM API fees plus $3-7/mo for the server.
Does n8n charge extra for AI agent workflows?
No. AI agent workflows use the same execution model as regular workflows. There’s no AI surcharge or separate pricing tier. You pay your normal n8n cost (Cloud plan or self-hosted server) plus whatever LLM API fees you incur from OpenAI, Anthropic, or other providers.
I wrote a more detailed comparison of every self-hosting option on Dev.to if you want the full breakdown with setup instructions for each method.
The bottom line: n8n Cloud is great if you want zero ops and your execution count is low. For everyone else, self-hosting saves $200-700/year and removes all execution limits.
Deploy n8n on InstaPods - 60 seconds, $3/mo
Related reading:
- n8n vs Zapier - detailed feature and pricing comparison
- n8n vs Make - which automation platform is cheaper at scale?
- Best Lightweight Server Monitoring Tools - if you’re self-hosting n8n, you’ll want monitoring too
- Deploy Without DevOps - how InstaPods makes self-hosting painless