n8n — Source-available automation you can self-host for free, with per-execution cloud pricing and a leading AI-agent builder.

Verdict at a glance

Overall rating: 4.5/5 ★★★★½
Standout: Own your automation — free self-hosting plus the most complete visual AI-agent stack.
Watch out: You need at least one technical person; self-host maintenance time can exceed the hosting fee.

The best value and control for technical teams and AI-agent builders. Budget engineering time if you self-host, and avoid cloud for high-frequency simple tasks.

Quick facts

• Ideal team: Technical teams with an engineer
• Learning curve: High
• Setup time: Hours-days (self-host)
• Support: Community (self-host); paid support on cloud/enterprise
• Key limit: Self-hosting carries real ops/maintenance burden
• Top integrations: 500+ integrations, Code nodes, AI/LangChain

Overview

n8n is a source-available automation platform you can self-host for free (Community Edition, unlimited executions) or run as managed cloud. It bills per workflow execution rather than per action — one run equals one execution regardless of node count — so deep multi-node workflows don't inflate cost.

It supports JavaScript/Python expressions in nodes and has become a leading platform for building AI agents: a dedicated AI Agent node handles tool calling, vector-store memory, and multi-step reasoning across LLMs like GPT-4 and Claude. It's the favorite of technical teams who want control, data residency, and predictable pricing.

Key features

• Self-hostable, free Community Edition (unlimited executions)
• Per-execution billing (one run = one execution)
• 500+ integrations
• Code nodes (JavaScript/Python expressions)
• AI Agent node + LangChain (tool calling, vector memory)
• Sub-workflows; unlimited workflows/users
• Git, SSO, multi-environment (Business)

Pricing (as of 2026-06-12)

Model: per-execution cloud OR free self-host · Free tier: yes (self-hosted Community Edition, unlimited) · Entry paid: Cloud Starter ~EUR24/mo

Annual billing is ~17% cheaper. Cloud model can be costly for high-frequency simple automations vs self-hosting.

⚠️ Self-hosting is free but carries real ops time (updates, backups, monitoring). Pricing changes often — confirm on the vendor site.

Ratings breakdown

• Ease Of Use: 3.0/5
• Features: 4.5/5
• Value: 5.0/5
• Support: 3.5/5

Pros

• Free self-hosting with the full feature set
• Per-execution pricing scales well for deep workflows
• Code-level flexibility (JS/Python)
• Leading visual AI-agent builder (GPT-4/Claude, memory, tools)

Cons

• Steeper setup than Make
• Self-hosting has real ops cost (~4 hrs/mo: updates, backups, monitoring)
• Cloud pricing punishes high-frequency simple automations
• Complex AI agents can need 15+ chained nodes and raw-JSON debugging
• Some cloud outages reported

Who it’s for

• Technical founders and AI builders
• Ops teams with at least one engineer
• Data-residency needs and AI-agent backends

Who it’s not for

• Solo non-technical users wanting the fastest path (Zapier)
• High-frequency simple automations on cloud pricing

Best use cases

• Self-hosted automation
• AI agent orchestration & LLM workflows
• Internal tooling

Integrations

500+ integrations, code nodes, AI/LangChain nodes, HTTP/API.

Alternatives

Make, Zapier, Activepieces.

Recent changes

Expanded AI Agent node with tool calling and vector-store memory.

Sources

• https://n8n.io/pricing/
• https://www.roborhythms.com/n8n-review/
• https://dev.to/nova_gg/n8n-review-2026-i-used-it-for-8-months-to-build-ai-agents-honest-verdict-kif

Compiled from public sources; last verified 2026-06-12. Pricing is subject to change — verify with the vendor. Ratings reflect editorial opinion.