n8n's pricing page shows a straightforward tiered structure for n8n Cloud, and a "free" self-hosted option that isn't actually free once you account for infrastructure and maintenance. Neither half of that picture is dishonest, but neither is complete on its own, and businesses evaluating n8n deserve the full real-cost picture before choosing between the two paths.
n8n Cloud: The Published Tiers
| Tier | Who it fits | What scales with price |
|---|---|---|
| Starter | Individuals and small teams beginning with automation | Execution volume and active workflow count |
| Pro | Growing teams with meaningful automation volume | Higher execution limits, more concurrent workflow executions |
| Enterprise | Organizations needing SSO, advanced permissions, and dedicated support | Custom pricing based on scale and specific enterprise requirements |
n8n Cloud pricing is structured around monthly workflow executions rather than a flat per-user seat price, which rewards businesses with many users triggering a moderate number of workflows over businesses with few users triggering an enormous volume, worth checking against your specific usage pattern rather than assuming pricing scales purely with team size.
Self-Hosted n8n: The Real Total Cost
| Cost component | Typical range | Notes |
|---|---|---|
| VPS/server infrastructure | ~$20-80/month | Scales with automation volume and complexity; most SMB workloads fit comfortably in this range |
| Initial setup and configuration | One-time, several hours to a few days | Docker deployment, database setup, security hardening |
| Ongoing maintenance | A few hours monthly | Security patches, n8n version updates, monitoring |
| Backup and disaster recovery | Minimal additional infrastructure cost | Genuinely necessary, frequently skipped by DIY setups |
The "free" in self-hosted n8n refers to the software license, not the total cost of running it reliably. For a business without existing DevOps capacity, this real cost is better captured as either the value of internal time spent on infrastructure management, or the cost of a managed n8n hosting service (PURIST or equivalent providers) that handles this layer.
The Crossover Point: When Self-Hosted Wins on Cost
| Monthly executions | n8n Cloud cost (approx.) | Self-hosted cost (infrastructure only) |
|---|---|---|
| 10,000 | Comparable to entry Cloud tier | Comparable, infrastructure minimums apply either way |
| 100,000 | Meaningfully higher Cloud tier required | Same infrastructure cost as 10,000 (no execution-based scaling) |
| 1,000,000+ | Enterprise-tier Cloud pricing | Same infrastructure cost, perhaps a modestly larger server |
Self-hosted n8n's cost is driven by infrastructure capacity, not execution count, which means the crossover point where self-hosting becomes dramatically cheaper than Cloud arrives faster than most businesses expect, often well before 100,000 monthly executions, depending on the specific Cloud tier pricing at the time.
Hidden Costs on Both Paths
On Cloud: exceeding your plan's execution limit typically triggers an upgrade requirement or throttling, worth monitoring proactively rather than discovering via a mid-month automation failure. On self-hosted: underestimating the maintenance time commitment is the most common hidden cost, an unmaintained self-hosted instance accumulates security and stability risk that eventually costs far more in incident response than the maintenance would have cost proactively.
Sources & Further Reading
Pricing structures reflect n8n's publicly listed plans as of publication; verify current tier pricing and execution limits directly on n8n's pricing page, as SaaS pricing changes periodically. See n8n's official pricing page for current rates. For the broader self-hosted-versus-cloud architecture decision beyond just cost, see n8n self-hosted vs cloud: the complete decision guide.
The 3-Year Total Cost of Ownership View
Pricing comparisons based on today's volume understate how the decision plays out as automation adoption grows within a business. Modeling growth from 20,000 to 200,000 monthly executions over three years:
| Path | Year 1 (20k executions/mo) | Year 3 (200k executions/mo) | 3-year total (est.) |
|---|---|---|---|
| n8n Cloud | Entry-to-mid tier pricing | Meaningfully higher tier required | Total scales substantially with volume growth |
| Self-hosted n8n | ~$20-40/mo infrastructure | ~$40-80/mo infrastructure (larger server for higher throughput) | Total stays roughly flat, scaling with server capacity, not execution count |
This is the same underlying dynamic covered in our n8n Cloud vs self-hosted decision framework: self-hosted's cost advantage compounds specifically as volume grows, which is why the right choice can genuinely change over a business's growth trajectory even if Cloud was the right starting point.
What "Fully Loaded" Self-Hosted Cost Actually Includes
A complete self-hosted cost model includes more than the raw server bill: the PostgreSQL database infrastructure (covered in our PostgreSQL setup guide), backup storage, and either internal engineering time or a managed-hosting fee for ongoing maintenance. A realistic fully-loaded monthly estimate for a properly maintained production self-hosted instance, including a managed hosting relationship, commonly runs $150-500/month depending on scale and support level, still typically well below equivalent Cloud tier pricing at meaningful execution volume, but meaningfully more than the bare "$20/month VPS" figure that circulates as the headline self-hosted cost.
Negotiating Enterprise Pricing
For organizations approaching n8n Cloud's Enterprise tier threshold, published pricing is typically a starting point for a sales conversation rather than a fixed rate; volume commitments, multi-year agreements, and specific feature requirements (SSO, dedicated support SLAs) are all commonly negotiable. Approach this the same way you would any enterprise SaaS negotiation, with your actual usage projections and specific must-have features documented before the conversation starts.
What to Model Before Choosing a Plan
- What is our realistic monthly execution volume today, and at 2x and 5x that volume?
- Does our workflow design minimize unnecessary executions (avoiding overly frequent polling triggers, for instance), or are we paying for inefficient workflow design rather than genuine automation need?
- Have we priced the fully-loaded self-hosted cost (infrastructure plus maintenance time or managed hosting fee), not just the bare server cost?
- At what specific execution volume does our own crossover point between Cloud and self-hosted actually fall, given current published pricing?
Sources & Further Reading
Pricing figures reflect n8n's publicly listed rates as of publication; SaaS pricing changes periodically, verify current tiers directly on n8n's official pricing page before budgeting.
For related reading, see n8n Cloud vs self-hosted: the real decision framework for the broader decision beyond pure cost, and n8n Docker: the production deployment guide for the self-hosted infrastructure this pricing model assumes.
Common Mistakes When Budgeting for n8n
Budgeting only for today's volume without a growth buffer. Automation adoption tends to accelerate once a team sees early wins; a budget with no headroom for volume growth gets outgrown faster than expected.
Comparing Cloud tier prices without checking what each tier actually includes. Higher tiers often unlock not just more executions but features (advanced permissions, priority support) that matter independently of raw volume; compare full feature sets, not just the headline execution limit.
Treating self-hosted infrastructure cost as the complete self-hosted cost. As covered throughout this guide, maintenance time or a managed hosting fee is a real, recurring cost that belongs in the comparison, not an afterthought discovered after committing to self-hosted.
A Realistic Scenario: A Mid-Year Tier Upgrade
A growing agency on n8n Cloud's entry tier saw execution volume grow 3x within six months as they onboarded new clients faster than planned, triggering an unplanned mid-cycle upgrade to a higher tier at a meaningfully higher monthly cost than budgeted. In their post-mortem, the team identified that they'd modeled their initial Cloud tier choice against current client count rather than their actual sales pipeline and onboarding velocity. For their next budget cycle, they modeled pricing against a 6-month-forward volume projection instead of current-state volume, avoiding a repeat of the same unplanned upgrade.
The 3-Year View: Total Cost Across a Realistic Growth Path
| Year | Realistic scenario | Approx. path |
|---|---|---|
| Year 1 | Starting automation, low volume, uncertain long-term needs | n8n Cloud entry tier, lowest setup friction |
| Year 2 | Volume and team automation literacy both growing | Either continued Cloud tier upgrades, or a planned migration evaluation |
| Year 3 | Mature automation practice, meaningful volume | Self-hosted (direct or managed) for most businesses that reach real scale, given the flat cost curve |
This progression isn't universal, a business with strict compliance needs or strong technical capacity from day one may reasonably start self-hosted immediately, but it reflects the most common realistic path PURIST sees across client engagements, and it's a reasonable default planning assumption for a business genuinely uncertain where to start.
A Complete Worked Budget Example
To make this guide's cost framework fully concrete, here's a realistic annual budget model for a 25-person company running a moderate automation practice, comparing both paths at their respective points on the pricing curve.
| Cost category | n8n Cloud path (annual) | Self-hosted path (annual) |
|---|---|---|
| Platform/infrastructure cost | Mid-tier Cloud plan, scaled for ~50,000 monthly executions | ~$40/month VPS + ~$25/month managed PostgreSQL |
| Setup/onboarding time | Minimal, near-immediate start | Meaningful one-time setup investment, 1-3 days of technical time |
| Ongoing maintenance | None beyond normal workflow building | A few hours monthly, or a managed hosting fee covering this |
| AI model usage (if using AI Agent nodes) | Same cost either way, billed by the model provider directly | Same cost either way, billed by the model provider directly |
| **Estimated annual total** | **Meaningfully higher, scaling with execution volume** | **Meaningfully lower, roughly flat regardless of execution volume within a wide range** |
This model illustrates the pattern covered throughout this guide without pretending to universal precision, your specific numbers depend on exact current pricing, your real execution volume, and your specific infrastructure choices, but the structural relationship, Cloud costs scale with volume while self-hosted costs scale with infrastructure capacity, holds consistently across the businesses PURIST has helped model this decision.
Where Pricing Transparency Falls Short Industry-Wide
n8n is more transparent about its pricing than many automation platform competitors, publishing actual tier pricing rather than requiring a sales call for basic rate information. Even so, the true self-hosted cost (infrastructure plus maintenance) is inherently harder to pin down with a single number than a published Cloud tier price, since it depends on choices you make (server sizing, managed versus DIY maintenance) rather than a fixed vendor rate. This is exactly why this guide emphasizes modeling your own specific situation rather than relying on any single published figure, including our own, as a precise prediction of your actual cost.
Frequently Asked Questions
Is self-hosted n8n really free?
The software license is free; running it reliably has a real infrastructure and maintenance cost, typically $20-80/month in server costs plus ongoing time investment, which is genuinely lower than Cloud pricing at real volume but not literally zero.
What counts as an "execution" in n8n Cloud's pricing model?
Generally, each time a workflow runs from its trigger to completion counts as one execution, regardless of how many nodes it contains; verify the current precise definition against n8n's documentation since billing definitions can have important edge cases (like sub-workflow executions) worth understanding before committing to a tier.
Can I start on n8n Cloud and migrate to self-hosted later?
Yes, and this is a common growth path, workflows built on Cloud can generally be exported and imported into a self-hosted instance, though verify version compatibility and any Cloud-specific features you may be using before planning a migration.
Does a managed n8n hosting service cost more than self-hosting myself?
Typically yes in raw infrastructure cost, but it eliminates the maintenance time and expertise requirement, which for a business without in-house DevOps capacity is usually a better trade than the cost difference, similar to the calculation covered in our n8n vs Make vs Zapier comparison regarding managed versus DIY infrastructure.
How do I estimate my expected monthly execution volume before choosing a plan?
Count your currently planned workflows, estimate how often each will trigger (daily, hourly, per-event), and multiply out a realistic monthly total, then add meaningful headroom for growth and workflows you haven't built yet, since underestimating volume is the most common cause of an unplanned mid-year plan upgrade.
Does n8n offer discounts for startups or nonprofits?
Check directly with n8n's sales team for current startup or nonprofit program availability, as these programs and their terms change over time and aren't always prominently advertised on the standard pricing page.
Is annual billing meaningfully cheaper than monthly on n8n Cloud?
Most SaaS platforms, n8n included, typically offer a discount (commonly in the 15-20% range) for annual commitment versus monthly billing; weigh this against the flexibility of staying monthly if you're still validating fit or expect volume to change significantly.
Can I mix Cloud and self-hosted, running some workflows on each?
Yes, some organizations do run a hybrid setup, for example Cloud for a business team's simpler automations and self-hosted for a technical team's higher-volume or compliance-sensitive workflows, accepting the added complexity of managing two environments for the specific benefits each provides.
Does n8n charge differently for workflows with AI Agent nodes versus standard workflows?
n8n's own platform pricing (Cloud tiers or self-hosted infrastructure cost) doesn't differentiate by node type; however, AI Agent nodes incur separate, additional cost from whichever AI model provider you're calling (Claude, OpenAI, etc.), billed directly by that provider based on token usage, a cost layer entirely separate from n8n's own pricing.
Is there a cost difference between triggering workflows via webhook versus polling?
Not in n8n's own pricing model directly, both count as executions; however, polling triggers on a short interval generate far more total executions over time than event-driven webhooks for the same underlying volume of real events, which indirectly affects cost on execution-based pricing tiers.
How do I budget for n8n cost growth as my business scales?
Build a simple model tying expected execution volume to your headcount or transaction volume growth projections, then check that projection against current tier pricing (Cloud) or infrastructure sizing (self-hosted) annually as part of normal budget planning, rather than treating automation cost as a fixed, one-time line item.
Does n8n Cloud pricing differ by region or currency?
Pricing is typically published in a primary currency (commonly USD or EUR) with regional payment processing; verify current regional pricing and any currency-conversion implications directly against n8n's pricing page for your specific billing location.
Are there additional costs for using n8n's AI Agent node beyond standard execution pricing?
The AI Agent node itself doesn't add a separate n8n platform fee beyond standard execution counting, but the underlying AI model calls it makes (to Claude, OpenAI, or another provider) are billed separately by that provider based on token usage, a real additional cost to budget for separately from n8n's own pricing.
Is it possible to get a custom enterprise quote below the published Enterprise tier starting price?
For genuinely large-scale or multi-year commitments, custom negotiated pricing below the nominal starting Enterprise price is often possible; approach this as a standard enterprise software negotiation with your specific volume and commitment terms as leverage.
Is it worth paying for n8n Enterprise features even below the volume that would technically require it?
Only if a specific Enterprise feature (SSO, advanced permissions) is a genuine current requirement, not a speculative future one; paying for Enterprise-tier features ahead of an actual need is a common way to overspend on automation infrastructure.
How often does n8n change its pricing structure?
Like most SaaS platforms, pricing structures and tier boundaries do change periodically; treat any specific figures in guides like this one, including our own, as directionally accurate rather than a guaranteed current rate, and verify directly before budgeting.
Does n8n publish case studies with real customer cost figures?
n8n publishes some customer case studies highlighting outcomes and use cases; specific detailed cost figures are less commonly disclosed publicly by name, verify directly with n8n if a specific comparable case study would help your own budgeting decision.
Is it possible to pilot n8n Cloud and self-hosted simultaneously before choosing?
Yes, running a small pilot on both in parallel, the same handful of test workflows built on each, gives a genuinely informed comparison of setup effort and day-to-day experience before committing budget and migration effort to either path exclusively.
What's the best way to track our actual n8n costs over time to catch surprises early?
Set a monthly calendar reminder to check actual execution volume against your current tier or infrastructure capacity, treating this as routine financial hygiene rather than something to discover only at renewal or when a limit is unexpectedly hit.
Does PURIST charge separately for the platform cost when managing a client's n8n instance?
Platform costs (Cloud subscription or self-hosted infrastructure) and the service fee for building, maintaining, and monitoring the automation are typically itemized separately in a client engagement, giving full visibility into what's a pass-through platform cost versus the value-added service itself.
Does n8n's pricing page show costs inclusive or exclusive of tax?
Published SaaS pricing is typically shown exclusive of applicable tax, which varies by jurisdiction; confirm your specific total cost including tax directly during checkout or with your billing contact rather than assuming the listed price is final.
Is there a community-maintained cost calculator for n8n pricing scenarios?
Some community members have shared informal spreadsheets or calculators for modeling Cloud versus self-hosted costs; treat any third-party calculator as a starting template to adapt with your own current pricing data rather than a guaranteed-accurate tool, since published rates change.
How often should this pricing guide itself be revisited for accuracy?
Treat published pricing guides, including this one, as a snapshot; revisit and verify directly against n8n's current pricing page at least annually or before any major budgeting decision, since SaaS pricing structures shift periodically.
To get an accurate cost projection for your specific automation volume and to decide between Cloud and self-hosted, book a free automation audit. We run both models for clients and can model your real cost before you commit to either path.
Tags
Purist Team
The PURIST editorial team covers automation, AI agents, and operations strategy for businesses scaling with n8n, Make, and Claude AI.
