The Agent ExaminerIndependent Review Authority
FrameworkNot yet hands-on tested

n8n

AI agents and workflows you can see and control.

What it is

n8n is a workflow automation platform for building AI agents and automations on a visual canvas with drop-to-code, connecting LLMs to your data across a large connector library. It runs on n8n Cloud or self-hosted.

At a glance

7/7 facts sourced
Pricing
Self-hosted Community Edition is free. Cloud (billed annually): Starter EUR20/mo (2.5K executions), Pro EUR50/mo (10K executions), Business EUR667/mo (40K, self-hosted), Enterprise custom — all with unlimited users, workflows, and integrations.source ↗
Hosting model
Two options — n8n Cloud (hosted) or self-hosted on your own infrastructure, including a free Community Edition on GitHub.source ↗
Memory & state
AI Agent nodes take memory sub-nodes: "Simple Memory" (in-instance, not persisted across sessions), "Postgres Chat Memory" for persistence, and a "Chat Memory Manager" for advanced management.source ↗
MCP support
Nativesource ↗
Integrations
400+ integrations per the GitHub repository (the homepage markets 500+); provider-agnostic LLM nodes via LangChain.source ↗
Model providers
LangChain-based; documented chat models include OpenAI, DeepSeek, Google Gemini, Groq, and Azure, and you can connect any model.source ↗
Open source
Nosource ↗
4.2of 5

Editorial scorecard

Overall 4.2 / 5

Our opinion — the same 0–5 rubric is applied to every platform, including Alfe. See the rubric.

Developer experience4/5
Pricing transparency & value4/5
Scalability4/5
Memory & state3/5
Integrations5/5
MCP nativeness5/5

Strengths

  • Hosted cloud and a free, self-hostable Community Edition.
  • Native MCP on both sides — an MCP Server Trigger node and an MCP Client node.
  • Large connector library plus provider-agnostic LLM nodes.

Trade-offs

  • Not OSI open source — it is fair-code / source-available under the Sustainable Use License v1.0, which restricts commercial redistribution.
  • Default "Simple Memory" isn't persisted across sessions; persistence needs an external store such as Postgres.
  • Cloud plans are metered by workflow executions.

Our take

Editorial opinion — not a hands-on test result

In our opinion, n8n is a compelling option when you want a visual, self-hostable automation platform with first-class MCP and a huge connector catalogue. The most important nuance is licensing: it is fair-code / source-available, not OSI open source, and the Sustainable Use License restricts commercial redistribution — read it before you embed or resell. For internal automation and self-hosting, it's a capable base.

Sponsored

Alfe is a managed agent-hosting platform. If you want an agent hosted and running without wiring infrastructure yourself, see Alfe → (or read its dossier).

FAQ

Is n8n open source?
Not in the OSI sense — it is fair-code / source-available under the Sustainable Use License v1.0. It is self-hostable for free, but commercial redistribution is restricted.
Does n8n support MCP?
Yes, natively — it provides both an MCP Server Trigger node and an MCP Client node.

Compare n8n

Sources

The Agent Examiner editorial · last updated 2026-07-01. Facts are drawn from each vendor’s own public materials; scores are our editorial opinion (methodology).

Product and company names are trademarks of their respective owners and are used here only to identify the products under discussion.