n8n Cloud vs Self-Hosted: The Complete Comparison Guide

You’re about to waste either thousands in subscription fees or hundreds of hours in infrastructure management. The choice between n8n Cloud and self-hosted determines whether you pay premium prices for convenience or invest serious time into running your own infrastructure.

Most comparisons gloss over the details that actually matter. They list features without explaining the real implications. They show pricing tiers without calculating what you’ll actually spend at scale. They mention “self-hosted is free” without acknowledging the hidden costs of time, expertise, and maintenance.

This guide goes deeper. We’ll analyze real costs at different execution volumes, compare the technical limitations that affect production workflows, and give you a clear framework for making the right decision for your specific situation.

The Stakes Are Higher Than You Think

Choosing the wrong deployment model creates problems that compound over time.

Pick Cloud when you should self-host? You’ll watch your monthly bill climb as your automation usage grows. That $24/month starter plan becomes $200+ when you scale, while a self-hosted instance would cost a fraction of that.

Pick self-hosted when you should use Cloud? You’ll spend dozens of hours learning Docker, debugging PostgreSQL connections, and troubleshooting SSL certificates. Hours that could have gone into building workflows or running your actual business.

Recent Changes That Matter

n8n has made significant pricing and feature changes that shift the calculation:

Removed active workflow limits from all plans (previously a major Cloud restriction)
Introduced a Business plan for self-hosted users who need SSO, Git workflows, and advanced features
Standardized execution-based pricing across Cloud plans
Added queue mode to Enterprise Cloud (previously self-hosted only)

These changes make both options more viable depending on your situation. The old advice of “always self-host for serious use” needs updating.

What You’ll Learn

The three n8n deployment options and what each actually includes
Real cost comparisons at 5K, 10K, 25K, 50K, and 100K monthly executions
Technical limits that affect production workflows (memory, timeouts, concurrency)
Data privacy and compliance considerations for regulated industries
Operational requirements and the true time investment of self-hosting
A decision framework based on your specific scenario
How to migrate between Cloud and self-hosted if you change your mind

The Three n8n Deployment Options Explained

Before comparing, you need to understand what you’re actually choosing between. n8n offers three distinct deployment models, each with different trade-offs.

n8n Cloud (Managed)

n8n Cloud is the fully managed option. You sign up, log in, and start building workflows immediately. n8n handles everything else: server infrastructure, database management, SSL certificates, automatic updates, and backups.

What you get:

Instant setup with no technical configuration
Automatic updates to the latest n8n version
Managed backups and disaster recovery
SSL and security handled for you
Official support included
EU-based data hosting (Frankfurt, Germany)

What you give up:

Execution limits based on your plan
Memory and CPU constraints
100-second webhook timeout (hard limit)
No queue mode on Starter/Pro plans
Less control over configuration and customization
Data stored on n8n’s infrastructure

Cloud works well for teams who want to focus entirely on building automations without any infrastructure overhead.

Self-Hosted Community Edition (Free)

The Community Edition is the open-source version of n8n that you run on your own infrastructure. It’s completely free with no execution limits, but you’re responsible for everything: deployment, security, backups, updates, and troubleshooting.

What you get:

No licensing costs whatsoever
Unlimited workflow executions
Full control over your data
Complete configuration flexibility
Queue mode for horizontal scaling
No vendor lock-in

What you give up:

You handle all infrastructure
Manual updates and security patches
Your own backup and monitoring setup
No official support (community forums only)
Requires DevOps knowledge

The Community Edition powers most serious n8n deployments. If you have the technical capability, it offers the best value at scale.

Self-Hosted with Business/Enterprise License

n8n also offers paid licenses for self-hosted deployments. These add enterprise features on top of the Community Edition while keeping your infrastructure under your control.

Business Plan features:

SAML/LDAP single sign-on (SSO)
Git-based workflow versioning
Environment management (dev/staging/prod)
Advanced role-based access control
Source-available code access
Official support

Enterprise Plan features:

Everything in Business
Custom integrations and SLAs
Dedicated success manager
Advanced audit logging

These plans suit organizations that need enterprise governance features but prefer to control their own infrastructure for compliance, security, or cost reasons.

Quick Comparison

Aspect	Cloud	Self-Hosted (Free)	Self-Hosted (Paid)
Cost	$24-$600+/month	Infrastructure only	License + infrastructure
Execution Limits	Plan-based	Unlimited	Unlimited
Setup Time	Minutes	Hours to days	Hours to days
Maintenance	None	Your responsibility	Your responsibility
Queue Mode	Enterprise only	Yes	Yes
Data Location	EU (Frankfurt)	Your choice	Your choice
Support	Included	Community only	Included

The Real Cost of Each Option

Pricing comparisons often miss the complete picture. Let’s break down what you’ll actually pay in different scenarios.

n8n Cloud Pricing Breakdown

n8n Cloud uses execution-based pricing. The current plans look like this:

Starter Plan: $24/month

2,500 executions included
320 MB RAM per execution
5 concurrent executions
Basic features

Pro Plan: $50/month

10,000 executions included
640 MB RAM per execution
10 concurrent executions
All integrations

Enterprise Plan: Custom pricing

Custom execution allocation
Up to 4 GB RAM per execution
Higher concurrency limits
Queue mode available
Dedicated support

The execution counting gotcha: n8n counts each node execution, not each workflow run. A workflow with 10 nodes that runs 100 times uses 1,000 executions, not 100. This surprises many users who expect workflow-level counting.

For a typical 15-node workflow running 200 times per day:

Daily executions: 15 × 200 = 3,000
Monthly executions: 3,000 × 30 = 90,000

That single workflow would exceed even the Pro plan’s 10,000-execution limit by day four.

Self-Hosted Cost Reality

Self-hosting appears “free” but has real costs:

Infrastructure costs:

Server Type	Monthly Cost	Suitable For
Basic VPS (1 CPU, 2GB RAM)	$5-12	Testing, light use
Standard VPS (2 CPU, 4GB RAM)	$20-40	Moderate production use
Production VPS (4 CPU, 8GB RAM)	$40-80	Heavy workloads
Managed PostgreSQL	$15-50	If you don’t want to manage DB

Hidden costs:

Initial setup time: 4-16 hours depending on experience
Ongoing maintenance: 2-4 hours monthly minimum
Learning curve: Varies significantly by background
Troubleshooting incidents: Unpredictable but real

If you value your time at $50/hour and spend 5 hours monthly on n8n maintenance, that’s $250/month in opportunity cost. For some teams, that makes Cloud the better deal. For others with existing DevOps capabilities, the marginal time cost is near zero.

Break-Even Analysis

Here’s where the math gets interesting. At what point does self-hosting become cheaper?

Monthly Executions	Cloud Cost	Self-Hosted Cost*	Winner
2,500	$24	$20-40	Cloud
5,000	$50+	$20-40	Self-hosted
10,000	$50+	$20-40	Self-hosted
25,000	$100+	$30-60	Self-hosted
50,000	$200+	$40-80	Self-hosted
100,000	$400+	$60-120	Self-hosted

*Self-hosted costs assume infrastructure only, not time investment.

Key insight: If you exceed 5,000 executions monthly and have any DevOps capability, self-hosting almost always makes financial sense. The savings grow dramatically at scale.

When Cloud still wins despite higher cost:

You have zero DevOps resources or interest
Speed to deploy is more valuable than cost savings
Your team’s time is better spent on core business activities
You need the managed experience for non-technical team members

Features, Limits, and What Actually Matters

Beyond pricing, technical limitations determine whether n8n fits your use case. Some limits are flexible, others are hard constraints.

Execution Limits and Performance

Cloud execution limits gate your usage based on your plan. Running out means workflows pause until the next billing cycle or you upgrade.

Self-hosted execution limits don’t exist at the software level. Your only constraints are the hardware you provision. Want to run 500,000 executions monthly? Provision enough CPU, memory, and configure queue mode with multiple workers.

Memory per execution matters for data-heavy workflows:

Scenario	Memory Needed	Cloud Starter (320MB)	Cloud Pro (640MB)	Self-Hosted
Simple API calls	< 100MB	Works	Works	Works
Moderate JSON processing	100-300MB	Borderline	Works	Works
Large file processing	500MB+	Fails	Borderline	Works
Heavy data transformations	1GB+	Fails	Fails	Works

If you process large spreadsheets, PDFs, or significant JSON payloads, Cloud memory limits become a real constraint.

Technical Limitations That Matter

100-Second Webhook Timeout (Cloud)

n8n Cloud enforces a 100-second timeout on webhook responses. If your workflow takes longer than 100 seconds to complete, the calling service receives a timeout error.

This affects:

Heavy data processing triggered by webhooks
Workflows that call slow external APIs
Batch operations triggered by incoming requests

Self-hosted n8n has no such limit. Configure timeouts as high as you need via environment variables.

Concurrent Execution Limits

Cloud plans limit how many workflows can run simultaneously:

Starter: 5 concurrent executions
Pro: 10 concurrent executions
Enterprise: Higher limits

Self-hosted concurrency is limited only by your server resources and configuration. With queue mode, you can run dozens or hundreds of concurrent executions across multiple workers.

Binary Data Storage

Cloud stores binary data (files, images) with plan-based limits. Heavy file processing workflows can hit these limits quickly.

Self-hosted uses your own storage. Connect S3, local disk, or any storage backend with no artificial limits.

Detailed Feature Comparison

Feature	Cloud Starter	Cloud Pro	Cloud Enterprise	Self-Hosted
Monthly executions	2,500	10,000	Custom	Unlimited
Memory per execution	320 MB	640 MB	Up to 4 GB	Your server
Concurrent executions	5	10	Custom	Your server
Webhook timeout	100 sec	100 sec	100 sec	Configurable
Queue mode	No	No	Yes	Yes
Custom nodes	Limited	Limited	Yes	Yes
External databases	No	No	Yes	Yes
SSO/SAML	No	No	Yes	Business+
Git versioning	No	No	Yes	Business+
Data location	EU only	EU only	EU only	Your choice

Scaling Capabilities

Cloud scaling means upgrading your plan. Need more executions? Pay more. Need more memory? Pay more. There’s no horizontal scaling option on Starter or Pro plans. Enterprise customers get queue mode, but it comes with enterprise pricing.

Self-hosted scaling follows a different path entirely. Our queue mode guide covers this in detail, but the basics:

Set up PostgreSQL (required for queue mode)
Add Redis as your queue backend
Deploy multiple n8n worker processes
Scale workers horizontally based on load

This architecture handles 10,000+ concurrent executions on modest hardware. Some self-hosted deployments process millions of executions monthly without breaking a sweat.

Data Sovereignty and Compliance Considerations

For many organizations, where data lives and how it’s handled isn’t optional. Regulatory requirements may dictate your deployment choice.

Where Does Your Data Live?

n8n Cloud stores all data on EU servers in Frankfurt, Germany. This includes:

Workflow definitions
Execution history and logs
Stored credentials (encrypted)
Binary data from file processing

You cannot choose a different region. If your compliance requirements mandate US-only, Asia-Pacific, or other specific data residency, Cloud isn’t an option.

Self-hosted stores data wherever you deploy. Run n8n on AWS us-east-1, a German data center, your own on-premises servers, or anywhere else. You control the data location completely.

Compliance Requirements

GDPR Considerations

n8n Cloud’s EU hosting helps with GDPR compliance for EU organizations. However, you still need to:

Ensure your workflow data handling aligns with GDPR principles
Implement proper data retention policies
Handle data subject requests appropriately

Self-hosted gives you full control over GDPR compliance, but the responsibility is entirely yours.

HIPAA Requirements

Healthcare organizations handling Protected Health Information (PHI) typically cannot use standard Cloud offerings. HIPAA requires:

Business Associate Agreements (BAAs)
Specific security controls
Audit trails and access logging

n8n Cloud Enterprise may offer BAAs (contact n8n directly). Self-hosted with proper security configuration is the common choice for HIPAA-regulated workflows.

SOC 2 and Other Certifications

n8n has been working on compliance certifications for Cloud. Check their current documentation for the latest status. Self-hosted compliance depends entirely on your infrastructure and practices.

Credential Security

Both deployment options encrypt credentials at rest, but the implementation differs.

Cloud manages encryption keys automatically. You trust n8n’s security practices and infrastructure.

Self-hosted requires you to set the N8N_ENCRYPTION_KEY environment variable. This key encrypts all credentials in the database. You’re responsible for:

Generating a strong key
Storing it securely
Using the same key across all n8n instances
Never losing or exposing it

For detailed credential security practices, see our credential management guide.

What It Actually Takes to Run Each Option

Theory aside, let’s talk about the practical work involved in each deployment model.

Cloud Operations (Almost Zero)

Running n8n Cloud requires minimal operational effort:

Sign up and verify your account
Start building workflows immediately
Monitor execution counts to stay within limits
Upgrade plans if you need more capacity

That’s it. No servers to maintain, no databases to back up, no security patches to apply. n8n handles infrastructure, and you focus on automation.

Support is included. When something goes wrong or you have questions, n8n’s support team helps. For Starter and Pro plans, this is typically email-based. Enterprise customers get dedicated support channels.

Self-Hosted Operations (The Real Work)

Self-hosting requires ongoing investment. Here’s what you’re signing up for:

Initial Setup (4-16 hours)

Provision a server (VPS, cloud instance, or on-premises)
Install Docker and Docker Compose
Configure PostgreSQL for production use
Set up a reverse proxy (nginx, Traefik, or Caddy)
Obtain and configure SSL certificates
Configure environment variables
Set up authentication
Test the deployment

Ongoing Maintenance (2-4 hours monthly)

Apply n8n updates (released regularly)
Monitor server health and resources
Review and rotate backups
Apply security patches to the host OS
Respond to any incidents or failures
Scale resources as usage grows

Incident Response (Variable)

When something breaks, you fix it. This might be:

Database connection issues
Memory exhaustion from heavy workflows
SSL certificate expiration
Server hardware failures
n8n bugs requiring workarounds

Our guide on common self-hosting mistakes covers the issues most likely to bite you.

The DevOps Reality Check

Before committing to self-hosted, honestly assess your team:

Questions to ask:

Does anyone on the team have Linux server experience?
Are you comfortable with Docker and container management?
Can you debug database connection issues?
Do you have a process for applying security updates?
What happens when n8n goes down at 2 AM?

If the answers make you uncomfortable, either invest in building that capability or consider Cloud. There’s no shame in paying for managed infrastructure. It’s a legitimate trade-off.

When to hire help:

If self-hosting makes sense financially but you lack the expertise, consider:

Our n8n self-hosted setup service for initial deployment
Our support and maintenance service for ongoing management
Consulting to architect your deployment properly

The goal is reliable automation, not proving you can manage servers.

Which Option Is Right for You?

Let’s cut through the complexity with clear recommendations based on common scenarios.

Choose n8n Cloud If…

You’re a solo founder or small team without DevOps resources. Your time is better spent building your business than managing infrastructure. The premium you pay for Cloud buys focus.

Your execution volume is moderate. If you’ll stay under 10,000 executions monthly, Pro plan costs are reasonable compared to the management overhead of self-hosting.

You need to start immediately. Cloud gets you building within minutes. Self-hosting takes hours to days. If time-to-value matters more than long-term cost, Cloud wins.

Non-technical team members need access. Cloud’s managed experience means no risk of someone accidentally breaking the server or database.

You’re prototyping or validating an idea. Invest in infrastructure after you’ve proven the concept works.

Choose Self-Hosted Community Edition If…

You exceed 10,000 executions monthly. The cost savings become significant and grow with scale. At 50,000+ executions, you’re saving hundreds monthly.

You have DevOps capability. Whether in-house or through partners, someone can manage the infrastructure without it becoming a time sink.

Data sovereignty is required. Your data must stay on your infrastructure, in specific regions, or under your direct control.

You need queue mode for scaling. Processing high volumes of concurrent workflows requires queue mode, which isn’t available on Cloud Starter or Pro.

Webhook timeouts are a problem. Long-running workflows triggered by webhooks need the configurable timeouts only self-hosting provides.

You’re building a business on n8n. Agencies and companies where n8n is core infrastructure benefit from the control and cost efficiency of self-hosting.

Choose Self-Hosted Business/Enterprise If…

You need SSO integration. Enterprise identity management with SAML or LDAP is a Business plan feature.

Git-based version control is required. Teams that need workflow versioning, branching, and environment promotion benefit from the Business plan’s Git integration.

Compliance requires advanced audit trails. Enterprise plans offer enhanced logging and audit capabilities.

You want official support with self-hosted. Community Edition has community support only. Paid plans include n8n’s support team.

Decision Table by Scenario

Your Situation	Recommendation
Solo founder, testing automation ideas	Cloud Starter
Small business, <5K monthly executions	Cloud Starter or Pro
Agency, 5-20K monthly executions	Cloud Pro or Self-hosted
Growing business, 20K+ executions	Self-hosted Community
Enterprise, 50K+ executions	Self-hosted with Enterprise license
Healthcare, finance, or other regulated industry	Self-hosted
Non-technical team, any volume	Cloud (appropriate tier)
Heavy file processing or large payloads	Self-hosted
Need queue mode for high concurrency	Self-hosted or Cloud Enterprise

Switching Between Cloud and Self-Hosted

Your decision isn’t permanent. You can migrate in either direction if your needs change.

Cloud to Self-Hosted Migration

Moving from Cloud to self-hosted is the more common direction as organizations scale.

Step 1: Export your workflows

In n8n Cloud, export each workflow as JSON (three-dot menu → Download). This captures the workflow structure but not credentials or execution history.

Step 2: Set up your self-hosted instance

Follow our Docker setup guide and PostgreSQL configuration guide to deploy n8n on your infrastructure.

Step 3: Import workflows

Import your JSON files into the self-hosted instance. Workflows appear inactive, ready for credential configuration.

Step 4: Reconfigure credentials

You cannot export credentials from Cloud (for security reasons). Create new credentials in your self-hosted instance for each service your workflows use. This is the most time-consuming step for complex deployments.

Step 5: Parallel running (recommended)

Run both Cloud and self-hosted simultaneously during transition:

Activate workflows one by one on self-hosted
Verify they work correctly
Deactivate the Cloud version
Monitor for issues
Cancel Cloud subscription once migration is complete

This approach minimizes risk and allows rollback if problems arise.

For comprehensive migration planning, check our Zapier to n8n migration guide, which covers similar principles.

Self-Hosted to Cloud Migration

Moving to Cloud is less common but sometimes makes sense:

Your team composition changed (lost DevOps expertise)
Management overhead exceeded expectations
You need guaranteed uptime without managing infrastructure

What you lose:

Unlimited executions (now plan-limited)
Queue mode (unless Enterprise)
Configurable timeouts (fixed at 100 seconds)
Data location choice (EU only)
Custom configuration options

Migration process:

Export workflows as JSON from self-hosted
Sign up for appropriate Cloud plan
Import workflows
Recreate credentials
Verify functionality
Decommission self-hosted infrastructure

Evaluate whether Cloud’s limits will work for your current usage before migrating. Hitting execution caps after migration creates problems.

Frequently Asked Questions

Can I start with Cloud and migrate to self-hosted later?

Yes. This is a common path. Start with Cloud to validate your automation use cases and learn n8n without infrastructure overhead. Once you’ve proven value and understand your execution volume, migrate to self-hosted for cost savings and control.

The migration involves exporting workflow JSON and recreating credentials on your self-hosted instance. Execution history doesn’t transfer, but workflow definitions do.

How do n8n Cloud execution limits work exactly?

n8n counts each node execution, not each workflow execution. A workflow with 10 nodes running once uses 10 executions. The same workflow running 100 times uses 1,000 executions.

This means a moderately complex workflow (15-20 nodes) running regularly can consume your monthly allocation quickly. A 15-node workflow running 200 times daily uses 90,000 executions monthly, far exceeding the Pro plan’s 10,000 limit.

Use n8n’s execution count tracking in the Cloud dashboard to monitor usage. Consider upgrading or migrating to self-hosted if you’re consistently hitting limits.

What’s the minimum server for self-hosting n8n?

For testing and light use, a basic VPS with 1 CPU and 2GB RAM ($5-12/month) works. For production with moderate traffic, we recommend 2 CPUs and 4GB RAM ($20-40/month). Heavy workloads or queue mode deployments need 4+ CPUs and 8GB+ RAM.

PostgreSQL is required for production. Either self-host it alongside n8n or use a managed database service. The database doesn’t need powerful hardware unless you retain extensive execution history.

See our Docker setup guide for specific configuration recommendations.

n8n Cloud stores data on EU servers in Frankfurt, Germany, which helps with GDPR compliance for EU organizations. However, GDPR compliance depends on how you use n8n, not just where it’s hosted.

You’re still responsible for:

Only processing data with proper legal basis
Implementing appropriate retention policies
Handling data subject requests
Ensuring your workflow logic respects privacy principles

For specific compliance requirements, consult the n8n documentation and consider legal advice for your situation.

Can I use queue mode on n8n Cloud?

Queue mode is only available on n8n Cloud Enterprise plans. Starter and Pro plans run all executions on a single process without horizontal scaling.

If you need queue mode for high-volume concurrent processing, your options are:

Upgrade to Cloud Enterprise (contact n8n for pricing)
Self-host and configure queue mode yourself

Our queue mode guide explains the architecture and setup for self-hosted deployments. Queue mode requires PostgreSQL and Redis but enables processing thousands of concurrent executions.

Making Your Decision

The right choice depends on your specific situation, not general best practices.

If infrastructure isn’t your strength and budget allows, Cloud removes complexity. You’ll pay more per execution but gain predictable operations and zero maintenance burden.

If you’re technically capable and cost-conscious, self-hosting delivers better value at scale. The initial investment pays off within months for high-volume deployments.

If you’re uncertain, start with Cloud. It’s easier to migrate to self-hosted later than to struggle with infrastructure you’re not ready to manage.

Whatever you choose, build your automations with production patterns from the start. Our workflow best practices guide covers error handling, monitoring, and maintainability, principles that apply regardless of deployment model.

Need help deciding or implementing? Our consulting service provides expert guidance on architecture decisions, and our self-hosted setup service handles the infrastructure work if you choose that path.

The goal is reliable automation that serves your business. Pick the deployment model that gets you there with the resources you have.