n8n: visual automation you can host yourself
n8n is an automation tool where you connect apps and AI agents by wiring blocks together on a canvas. It suits freelancers and teams who are comfortable with a bit of tech, especially if you want your data to stay on your own server.
Data as of October 22, 2026
Key figures
| Developer | n8n GmbH |
|---|---|
| Type | Low-code |
| Languages | JavaScript, TypeScript, Python |
| Open source | No |
| License | fair-code |
| First release | 2019 |
| Suggested level | Intermediate |
| Use cases | automation, workflows, RAG, chatbot |
| Self-hostable | Yes |
| Business model | Freemium |
| Cloud offering | Yes |
| Multi-agent | Yes |
| MCP support | Yes |
| Bring your own model | Yes |
| Official site | Official site · Documentation · Repository |
GitHub
| Repository | n8n-io/n8n |
|---|---|
| GitHub stars | 208,327 |
| Forks | 61,216 |
| Latest code activity | October 22, 2026 |
| Latest release | v1: :rocket: Release 1.123.86 (#41364) (October 22, 2026) |
GitHub activity checked on October 22, 2026.
What you can actually automate with n8n
n8n (pronounced “n-eight-n”, short for “nodemation”) is developed by n8n GmbH, and its GitHub repository dates back to 2019. A workflow starts from a trigger (an incoming email, a form submission, a schedule, a web call) and then runs a chain of blocks, called nodes, that read, transform and send data between apps. That is classic automation, with AI building blocks on top.
Typical examples: routing inbound requests and creating a CRM record, summarizing messages every morning, enriching a list of leads, or feeding a knowledge base for a chatbot. The AI Agent node turns a step into an AI agent that decides which tools to call; a main agent can even hand work to specialized sub-agents. n8n can also connect to MCP servers or expose its own workflows as an MCP server. When the built-in nodes are not enough, a Code node runs JavaScript or Python.
License and self-hosting: read this first
n8n calls itself “fair-code”: the source is visible and modifiable, but the Sustainable Use License restricts use to internal business, personal or non-commercial purposes. You can run as many free Community instances as you like, including to automate work for clients. White-labeling n8n, or letting outside users build their own workflows through your product, requires a commercial agreement. Some enterprise features (files marked “.ee”) fall under a separate license.
Self-hosting is the big draw for privacy-minded teams: data flows through your own server. The trade-off is that updates, backups and security are on you.
Cost and who it fits
As of October 22, 2026, the official pricing page lists, in euros and billed annually: Starter at €20 per month, Pro at €50 per month, Business at €667 per month (self-hosted only) and Enterprise on request. Pricing is based on workflow executions regardless of complexity, with unlimited users and workflows. AI model usage is billed separately by the provider; the agent cost calculator helps you estimate it.
- Freelancer: the Starter cloud plan or a small self-hosted instance is enough to begin.
- Small business: Pro, or self-hosting if someone can maintain the server.
- Technical team: n8n fits well thanks to code nodes, webhooks and its API.
Strengths, limits and alternatives
Strengths: a lot of flexibility, the option to host it yourself, and no per-step billing. Limits: a steeper learning curve than consumer tools, a license you must read carefully if you build a product on top, and agents that need testing before you let them take sensitive actions (the Can I trust an agent with this? tool helps you think it through).
If you want something simpler and fully hosted, look at Make or Zapier Agents. For AI apps built around a chatbot or search over your own documents, Dify and Langflow are more specialized.
Frameworks of the same type
- Dify · chatbot, RAG, workflows
- Botpress · chatbot, RAG, automation
- Microsoft Copilot Studio · chatbot, workflows, automation
- Langflow · RAG, multi-agent, workflows
- Gemini Enterprise Agent Platform (ex-Vertex AI Agent Builder) · multi-agent, RAG, workflows
- Relevance AI · multi-agent, automation, workflows
Frequently asked questions
Is n8n open source?
Not in the strict sense. The code is public, but it is released under the Sustainable Use License, a fair-code license n8n created in 2022. It allows free internal, personal or non-commercial use, but forbids, for example, reselling n8n as a service where your customers build their own workflows.
How much does n8n cost?
The self-hosted Community edition is free. As of October 22, 2026, n8n's cloud starts at €20 per month (Starter plan, billed annually, 2,500 workflow executions per month), then €50 per month for Pro. The official pricing page shows prices in euros.
Can I build automations for my clients with n8n?
Yes, according to the official license FAQ: you can build automations for clients on your own instance and charge for setup and maintenance, as long as clients cannot create or edit the workflows themselves. If you are unsure, n8n asks you to contact its licensing team.
Does n8n work with ChatGPT, Claude or local models?
n8n is not tied to one model provider: its AI nodes connect to several providers using your own API key. You then pay for model usage directly to that provider, on top of any n8n plan.
Data sources
- n8n.io
- github.com
- github.com
- docs.n8n.io
- docs.n8n.io
- GitHub — Repository metadata: GitHub
Data as of October 22, 2026
Report an error
Does a figure, a date or a fact look wrong? Tell us: every report is checked.