Loading directory…
| Tier | Price | Includes |
|---|---|---|
Community Edition | Free | Unlimited workflows and executions; self-hosted; source-available under Sustainable Use License; Docker/Kubernetes/npm install |
Business | Self-hosted. €667/mo billed annually. 40K executions/mo, SSO/SAML/LDAP, environments (dev/staging/prod), Git version control, queue mode scaling. Overage: €4,000 per 300K extra executions. | — |
Pro | n8n Cloud. €50/mo billed annually (€60 monthly). 10K executions/mo, 20 concurrent, 3 shared projects, admin roles, global variables, workflow history, execution search, 150 AI Builder credits. | — |
Starter | n8n Cloud. €20/mo billed annually (€24 monthly). Includes 2.5K workflow executions/mo, 5 concurrent executions, 1 shared project, unlimited users, 50 AI Workflow Builder credits. | — |
Enterprise | Contact sales | — |
What it does n8n is a self-hostable workflow automation platform with a visual node editor, 500+ integrations, and embedded AI agent nodes. You wire triggers (webhook, cron, queue) to actions (HTTP, database, SaaS API, LLM call) and optionally drop into JS/Python code nodes for custom logic.
Who it's for Platform, SRE, and internal-tools teams who want Zapier-class integration breadth without the per-task Zapier bill, and who can self-host. Also adopted by AI engineers building agentic flows that need real integrations (not just chat).
How platform engineers use it Self-host n8n on Kubernetes or Docker as the automation backbone for internal ops. Common patterns: enrich PagerDuty incidents with Datadog context and post to Slack; route GitHub webhook events through approval flows; pull Terraform Cloud run notifications into Linear; build ChatOps bots that call internal APIs; orchestrate AI agents that read from vector stores and write to Jira. Queue mode (Business+) lets you scale workers horizontally for high-throughput jobs.
Strengths
Limitations
AI maturity AI-enhanced rather than AI-native. The core product is workflow automation; AI agent nodes, vector store integrations, and the AI Workflow Builder are recent additions layered on top of a mature integration platform. The AI nodes are real (not marketing) and useful for orchestrating LLM calls inside broader workflows, but n8n's value remains the integration graph, not the model itself.
