Skip to main content
agent platform

n8n

Open-source workflow automation with first-class AI agent nodes.

Visit n8n Verified May 2026 how we review

What it is

n8n started as a Zapier alternative and evolved into a serious AI orchestration layer. You can self-host, wire any model into a node, and build deterministic + agentic flows side by side.

Notes from using it

n8n is what you reach for when you need real automation, can write some JS, and care about owning your data. Self-hosting on a $20/mo VPS gives you what teams pay $1,000+/mo for in equivalent Zapier usage. The trade-off is operational ownership — when n8n breaks at 2am, you're the one fixing it.

The AI nodes are powerful but under-documented compared to n8n's older deterministic nodes. Teams shipping AI-driven flows on n8n typically learn through community examples and discord, not official docs. That gap is closing but still real.

The natural failure mode: visual workflows grow until they're impossible to read. The escape hatch — JS function nodes — is what makes complex flows tenable, but it also means n8n at scale starts looking like a code project with extra steps. For pure no-code operators, Lindy or Zapier Agents is a cleaner fit. For technical teams that want self-hosting and full control, n8n is the strongest option in the category.

Where it shines

  • Self-hostable — keep data in your VPC.
  • Huge node library for SaaS integrations.

Where it falls down

  • AI features are powerful but underdocumented.
  • Complex workflows demand JS familiarity.

Best fit for

If you're trying to put AI behind any of these functions, n8n is worth a look:

Review changelog

What's changed since we first published this review. Newest first.

  • Initial review published. Pricing, positioning, and capability claims verified against n8n's docs and pricing page.

Head to head

n8n compared

Direct comparisons with the closest alternatives.

From the blog

Field reports mentioning n8n

Jun 3, 2026

Agent-to-Agent (A2A): the standard letting AI agents work together

What Google's A2A protocol is, how it pairs with MCP, and why multi-agent interoperability finally matters for operators building real stacks.

Jun 5, 2026

Glean vs Dust: Enterprise AI Search and Assistants Compared

Glean and Dust both promise work-grounded AI, but serve different operators. Here's how connectors, grounding quality, and pricing stack up.

May 27, 2026

The hidden architecture of production AI agents: what separates working deployments from demos

An operator's read on the engineering layers that actually determine whether an AI agent ships and holds up — CLAUDE.md investment, MCP tool surfaces, eval harnesses, cost controls, escalation patterns, observability, and the failure modes that don't show up until the agent is live.

Jun 4, 2026

Prompt caching: the AI cost lever most teams ignore

Prompt caching cuts token costs by up to 90% on repeated context. Here's how it works, which providers support it, and the gotchas operators miss.

Jun 6, 2026

Manus and the general-agent push: what a do-anything AI really automates

Manus, OpenAI Operator, and the general-agent wave promise to do everything. Here's what they actually automate—and where operators hit the wall.

Jun 12, 2026

How to Run an AI Vendor Pilot That Tells You Something Real

A practical playbook for operators: how to set success metrics, spot demo-ware, and build escape hatches into every AI vendor pilot.

Jun 13, 2026

Structured outputs: getting reliable JSON out of LLMs in production

How to use native structured outputs, tool-use schemas, Pydantic validation, and retry patterns to get schema-compliant JSON from LLMs reliably.

Jun 15, 2026

AI knowledge tools: turning company docs into something agents can use

Glean, Dust, Sana and their class of tools are redefining enterprise search. Here's what grounding, permissions, and RAG mean for operators deploying agents.

Jun 21, 2026

What an MCP Server Actually Is (and Why Operators Should Care)

Plain-English explainer of Model Context Protocol servers, clients, and hosts — and why every operator wiring AI agents to real tools needs to understand MCP now.

Jun 22, 2026

Clay vs Apollo for GTM Data: Enrichment Engine vs Data Provider

Clay and Apollo aren't competing tools—they're different layers of the same stack. Here's how operators think about the divide and when to run both.

Jun 24, 2026

Make vs n8n for AI Automation: Which Workflow Engine for Operators

Make vs n8n compared on AI agent capabilities, hosting flexibility, and true cost of ownership—so operators can pick the right engine.

n8n — frequently asked questions

What is n8n?

n8n started as a Zapier alternative and evolved into a serious AI orchestration layer. You can self-host, wire any model into a node, and build deterministic + agentic flows side by side.

How much does n8n cost?

Self-host free; Cloud Starter $24/mo. (Pricing verified as of May 2026.)

Who should use n8n?

n8n is a strong fit when: Self-hostable — keep data in your VPC. Huge node library for SaaS integrations.

What are the limitations of n8n?

Where it falls down: AI features are powerful but underdocumented. Complex workflows demand JS familiarity.

Is n8n both hosted and self-hostable?

n8n is both hosted and self-hostable. You can pick: use the hosted version for speed-of-deployment, or self-host for full data control.

What are the best alternatives to n8n?

The closest alternatives are , , . See our full n8n alternatives page for the curated list with trade-offs.