n8n is a workflow automation platform built for developers, IT teams, and technical operators who need to connect apps, automate business processes, and build AI agents without being boxed in by rigid no-code tools. It combines a visual, node-based canvas with the option to drop into JavaScript/Python code at any step, and it can be self-hosted or run on n8n's cloud infrastructure. The platform is aimed at teams doing IT operations, security operations, sales automation, backend prototyping, and RAG or multi-agent AI systems who need traceability and control over every automated step. n8n is fair-code licensed and has an active open-source presence on GitHub with a large community.
What it does
n8n lets users design automation workflows visually on a canvas, connecting triggers and actions across more than 500 integrations, while still allowing custom code for edge cases that pre-built nodes don't cover. It's positioned as an AI workflow automation platform: it can plug large language models and AI agents into a company's own data, let users inspect each reasoning step an agent takes, and keep a human in the loop before actions execute. Typical use cases described on the site include onboarding new employees, enriching security incident tickets, converting natural language into API calls, generating customer insights from reviews, building RAG pipelines, and embedding n8n's automation engine inside other products via its API.
Key capabilities
- Visual + code hybrid workflows: Build automations on a drag-and-drop canvas, then switch to custom JavaScript or API calls whenever a pre-built node isn't enough.
- 500+ integrations and custom API connections: Pre-built nodes cover common SaaS apps, with the ability to connect to virtually any service that exposes an API.
- Traceable AI agents: Multi-agent and RAG setups run on a canvas where every decision step is visible and inspectable, rather than hidden inside a black-box agent.
- Model flexibility: Swap between cloud-based and offline/local AI models without rebuilding the underlying workflow logic.
- MCP support: Build and update n8n workflows directly from inside compatible AI applications using the Model Context Protocol.
- Flexible deployment: Run workflows on n8n's hosted cloud service or self-host on private infrastructure for stricter data control, with enterprise options for SSO, SAML, LDAP, and version control via Git.
Pricing
n8n uses a tiered subscription model based on monthly workflow executions rather than user seats — all plans include unlimited users and workflows plus access to every integration. Confirmed plans (billed annually) include a Starter plan at €20/month for 2,500 executions with one shared project and 5 concurrent executions; a Pro plan at €50/month for 10,000 executions with 3 shared projects and admin roles; a Business plan at €667/month (self-hosted only) for 40,000 executions with SSO/SAML/LDAP and Git-based version control; and a custom-priced Enterprise plan with unlimited shared projects and dedicated support. Free trials are available on the paid tiers with no credit card required, and the open-source version can be self-hosted at no cost under n8n's fair-code license. Pricing page: View pricing
Editorial review
n8n's core strength is the hybrid model: it gives no-code-style speed for straightforward integrations while letting developers write custom code exactly where the pre-built nodes fall short, rather than forcing an all-or-nothing choice. Its execution-based (not per-user) pricing is favorable for teams with many collaborators but moderate workflow volume, and the self-hosting option is a genuine differentiator for organizations with data residency or compliance concerns that rule out fully cloud-based competitors. The AI-agent tooling — visible reasoning steps, human-in-the-loop checkpoints, and MCP integration — targets a real gap in agent frameworks that tend to obscure decision logic.
The trade-offs are also real: workflow-execution pricing can become unpredictable for high-volume automations, and the self-hosted Business tier requires infrastructure management that pure-SaaS alternatives avoid. The platform's power comes with a steeper learning curve than simple consumer automation tools like Zapier or Make, since getting full value often means comfort with JSON, APIs, or scripting. Best-fit users are technical teams — developers, IT/security operators, and platform engineers — who want fine-grained control over automation logic and AI agent behavior, rather than business users seeking a purely drag-and-drop experience. Teams evaluating n8n should confirm current execution limits and AI credit allowances directly on the pricing page, as these figures change with plan updates.
