n8n Pricing Explained: Cloud vs Self-Hosted for Startups (2026)

For most early-stage startups, n8n Cloud Starter (�24/mo) is the correct first choice. If you run more than 5,000 executions per month and have Docker experience, self-hosting on a �5-20/mo VPS saves �500-700/year. If you qualify for n8n’s startup program (under 20 employees, under �5M funding), the Business plan drops to �333/mo with 40,000 executions the best value tier available. Enterprise teams needing SSO and unlimited executions pay �2,000-3,000+/mo with custom contracts.

n8n Pricing at a Glance: Comparison Table

Plan	Monthly (Annual)	Executions	Concurrent	Key Features	Best For
Community Edition	Free	Unlimited	Hardware-limited	All 500+ integrations, unlimited users, core features	Technical solo builders, cost-sensitive teams
Starter	�24/mo (�20/mo)	2,500	5	1 shared project, 50 AI credits, forum support	Freelancers, early MVP testing
Pro	�60/mo (�50/mo)	10,000	20	3 shared projects, admin roles, workflow history, execution search, 150 AI credits, 7-day insights	Growing startups, small teams in production
Business	�800/mo (�667/mo)	40,000	200+	6 shared projects, SSO/SAML/LDAP, Git version control, environments, 30-day insights, scaling options	Mid-size companies, teams needing governance
Startup Plan	�333/mo (annual)	40,000	200+	Full Business features, 50% discount	Startups under 20 employees, under �5M funding
Enterprise	Custom (~�2,000+/mo)	Custom	200+	Unlimited projects, 365-day insights, 1000 AI credits, log streaming, external secrets, dedicated SLA support	Organizations with compliance/audit requirements

All plans include unlimited users, unlimited active workflows, and the full integration catalog. Prices verified against n8n.io/pricing as of May 2026. Annual billing saves 17%.

“n8n charges per execution one complete workflow run. Not per step, not per node. A 50-node workflow and a 3-node workflow each cost exactly one execution. This is n8n’s single biggest pricing advantage over task-based platforms like Zapier.”

How n8n Pricing Works: Bolded Definitions

Execution: One complete run of a workflow from trigger to finish, regardless of how many nodes, branches, or data transformations it contains. A workflow triggering on webhook, enriching data through 15 nodes, branching into 3 conditional paths, and writing to 4 different destinations = 1 execution.

Concurrent executions: How many workflows can run simultaneously. Starter caps at 5 if a 6th workflow triggers while 5 are already running, it queues. Business and Enterprise support 200+ concurrent runs.

Shared projects: Team collaboration buckets. Starter gives you 1 shared project (one shared workspace), Pro gives 3, Business gives 6, Enterprise is unlimited. This controls how many separate groups of workflows different teams can co-own.

Queue Mode: A self-hosted architecture using Redis that distributes incoming workflow executions across multiple worker instances. Essential for high-volume deployments processing thousands of executions per hour.

AI Workflow Builder credits: Metered tokens that power n8n’s built-in AI assistant (the tool that generates workflow JSON from natural language prompts). These are not your own AI/LLM API costs those you pay your provider directly.

Cloud vs Self-Hosted: The Real Cost Breakdown

1. n8n Cloud: What You Actually Pay

n8n Cloud is hosted in Frankfurt, Germany (EU data residency). You sign up, build workflows, and n8n handles everything below the application layer.

Starter (�24/mo): Covers 2,500 executions roughly 83 per day. One webhook-triggered workflow that fires 3-4 times per hour fits comfortably. Sufficient for lead notifications, Slack alerts, and simple CRM updates. Five concurrent executions means you can’t run complex parallel processing.

Pro (�60/mo): Covers 10,000 executions roughly 333 per day. Adds workflow history, execution search, admin roles, and global variables. This is the realistic production tier for startups running 10-20 workflows of moderate complexity. The 20 concurrent executions handle reasonable parallel loads.

Business (�800/mo): Covers 40,000 executions roughly 1,333 per day. The jump from Pro to Business is steep (13x cost for 4x executions), but the added value is SSO/SAML/LDAP authentication, Git-based version control, multi-environment support (dev/staging/prod), and 200+ concurrent executions. Overage pricing: �4,000 per additional 300,000 executions.

Enterprise (custom): Reported pricing starts around �2,000-3,000/month for organizations needing log streaming to external services (Datadog, etc.), external secret store integration, 365-day data retention, audit logging, and dedicated support with SLAs.

2. Self-Hosted n8n: Infrastructure Costs

The n8n Community Edition is free software unlimited executions, unlimited workflows, every integration. You only pay for the infrastructure it runs on.

Hosting Method	Monthly Cost	Maintenance	Executions
One-click managed (InstaPods, PikaPods)	�3-7	Zero (platform handles updates, SSL, backups)	Unlimited
Budget VPS (Hetzner CX22)	�4.50	1-2 hrs/mo (Docker, OS updates, SSL renewal)	Unlimited
Production VPS (4GB RAM, 2 vCPU)	�12-24	2-5 hrs/mo (full sysadmin duties)	Unlimited
Container platform (Railway, Northflank)	�5-50	Minimal (platform handles infrastructure)	Unlimited
Kubernetes (AWS/GCP/Azure)	�70-200	Significant (dedicated DevOps)	Unlimited

The hidden cost of self-hosting is engineering attention. At a fully-loaded engineer cost of �75-150/hour, 5 hours of monthly maintenance adds �375-750 in implicit cost pushing self-hosting past Cloud Business pricing even before infrastructure.

3. Execution Math: When Self-Hosting Wins

A typical early-stage startup running 4-6 workflows:

Slack notification bot: webhook-triggered, ~200/day
CRM sync: runs every 30 minutes, ~48/day
Lead enrichment: webhook-triggered, ~30/day
Weekly report: ~4/month
Onboarding automation: ~10/day

That’s ~8,640 executions/month nearly 3.5x the Starter limit, and 86% of the Pro limit. On n8n Cloud, you need Pro at �60/mo. On a one-click managed platform, the same workload costs �3-7/mo with no execution caps. Annual savings: �636-684/year.

For seed-stage startups running 20+ workflows with webhook-heavy triggers, 30,000-50,000 executions/month is common. Cloud Business at �800/mo vs self-hosted at �20-50/mo infrastructure = �9,000-9,360/year saved. That’s real headcount budget.

Decision Checklist: 7 Questions

How many executions per month? Under 2,500 Cloud Starter wins. 2,500-10,000 Cloud Pro is competitive. Over 10,000 self-hosting dominates on pure cost.
Do you have Docker/Linux expertise on the team? If yes, self-hosting is viable. If no, Cloud eliminates an entire category of operational risk.
Do you need SSO (SAML/LDAP)? Self-hosted Community Edition does not include SSO. You need Cloud Business (�800/mo), the Startup Plan (�333/mo), or self-hosted Business with a license key.
Does data need to stay in a private network? Cloud instances are in Frankfurt, Germany. If you need data on a specific VPC or behind a VPN, self-hosting is the only option.
Are workflows customer-facing or revenue-critical? If yes, treat n8n as production infrastructure. Either pay for Cloud with guaranteed uptime or invest properly in self-hosted monitoring, backups, and alerting.
Do you qualify for the startup discount? Under 20 employees and under �5M total funding = 50% off Business (�333/mo vs �800/mo). This narrows the gap significantly.
What’s your team’s opportunity cost? Every hour spent debugging Docker is an hour not spent on product. If your engineers cost �100/hr, self-hosting must save more than �100 per hour of maintenance to break even.

When n8n Cloud Is the Right Choice

You want to go from signup to working automation in 30 minutes
Nobody on the team wants to think about SSL certificates, Docker updates, or PostgreSQL vacuuming
Execution volume is predictable and under 10,000/month
You want managed upgrades with zero downtime risk
You qualify for the startup discount (�333/mo for Business features)
Your workflows use standard integrations without custom binary dependencies

When Self-Hosting Is the Right Choice

Execution volume exceeds 10,000/month and the cost gap is material
You need private network access to internal databases, ERPs, or APIs
Data sovereignty or regulatory compliance requires data to never leave your infrastructure
You need custom system libraries (FFmpeg, headless browsers, specific Python packages)
You run local AI models (Ollama, Llama) and need zero-latency inference without data egress
You already have Kubernetes, Docker, or similar infrastructure and monitoring in production
You’re building AI-native workflows that connect to local LLMs for zero-latency inference

The Startup Plan: n8n’s Best-Kept Secret

n8n offers a dedicated startup program: 50% off the Business plan, bringing it from �800/mo to �333/mo (billed annually) for 40,000 executions, SSO, environments, and Git version control.

Eligibility requirements (verified May 2026):

Fewer than 20 employees
Less than �5 million in total funding
Available for one year minimum (renewable if criteria still met)
Self-hosted deployment only (Business plan is self-hosted; Cloud tiers remain separate)

For a seed-stage startup that needs SSO for SOC2 compliance and Git for workflow versioning, this is the strongest value tier in the entire n8n pricing ecosystem.

Self-Hosted Business Plan: The Controversial Middle Ground

In 2026/2026, n8n introduced a self-hosted Business plan a license key that adds SSO, Git integration, environments, and scaling options to a self-hosted instance, but charges per-execution fees.

Key facts:

Base allocation of 40,000 executions included
Overage: �4,000 per 300,000 additional executions (roughly �0.013 per execution)
Failed workflows still count as executions
License key must ping n8n’s license server daily

This tier created community controversy because it introduced per-execution billing to self-hosted deployments something that didn’t exist before. For high-volume users (500,000+ executions/month), the math shifts: Community Edition on a �80 server may beat a licensed Business plan with overage fees. The Business plan becomes worthwhile only when you specifically need SSO or Git and your volume is moderate.

Hidden Costs Checklist

Self-Hosting Hidden Costs

PostgreSQL management: Vacuuming, connection pooling, backup verification ongoing attention required
SSL/TLS: Let’s Encrypt automates renewal but must be configured and monitored
Disk space: Docker images, WAL files, and execution logs accumulate automated pruning is essential
2 AM incidents: When disk fills up on a weekend and webhooks silently fail, that’s your problem
Update discipline: Breaking changes in n8n releases must be tested against your workflows before deploying

Cloud Hidden Costs

Overage charges: Business plan: �4,000 per 300,000 extra executions
AI credit limits: 50 on Starter, 150 on Pro, 1,000 on Enterprise complex AI agent generation burns through them
Data egress: Your data leaves your infrastructure for n8n’s EU servers a compliance consideration
Pricing evolution: n8n has restructured tiers twice in 18 months cloud pricing is not static

FAQ

Is n8n really free?

Yes. The self-hosted Community Edition is 100% free unlimited executions, unlimited workflows, every integration. You pay only for the server it runs on (as low as �3/month on one-click managed platforms).

What’s the cheapest way to use n8n?

For technical users comfortable with Docker: self-host on a �4-5/month VPS (Hetzner, DigitalOcean) or a �3-7/month one-click managed platform (InstaPods, PikaPods). For non-technical users: Cloud Starter at �24/month eliminates infrastructure risk entirely.

How does n8n compare to Zapier on cost?

n8n charges per execution (one complete workflow run). Zapier charges per task (every step counts). A 10-step workflow running 1,000 times costs 1,000 executions on n8n (�24/mo Starter covers this) but 10,000 tasks on Zapier (requires a much higher plan). At 50,000 executions with 15-step workflows, n8n can be 10-15x cheaper.

What happens when I exceed my execution limit?

On Cloud Starter and Pro: workflows stop running until the next billing cycle (no automatic overage billing). On Cloud Business: n8n sales reaches out, and overages are invoiced. On self-hosted Community Edition: no limits exist.

Does n8n charge per user?

No. Every plan includes unlimited users. This is a significant advantage over tools that charge per-seat your entire team can access the platform without incremental cost.

Where are n8n Cloud servers located? Frankfurt, Germany (EU). This provides GDPR-compliant data residency. If you need data in other regions, you must self-host.

Can I migrate between Cloud and self-hosted?

Yes. Export workflows as JSON from the UI or CLI. Credentials cannot be exported (security measure) you must re-authenticate integrations on the new instance. Update all webhook URLs in third-party services to point to your new domain. Migrate low-risk workflows first, validate behavior, then move critical workflows with rollback plans.

n8n Pricing Explained: Cloud vs Self-Hosted for Startups (2026)