n8n
Visit n8n →About n8n
n8n is a workflow automation tool that lets you wire together apps, databases and APIs through a visual editor, with the option to drop into code when you need it. It runs self-hosted or in the cloud.
The product has moved to the 2.x release line, with n8n@2.29.8 as the current stable build (released July 8, 2026) and 2.30.x in pre-release. The big 2026 change is that AI and agents are now front and center: the AI Agent node was rebuilt with structured tool calling for Claude, GPT-4o, Gemini, Mistral, Groq and OpenAI-compatible models, four memory backends (in-memory, Redis, Postgres, Motorhead) and a ReAct execution mode. The flat builder was replaced by a spatial Canvas UI introduced in v1.30.
Beyond AI, recent versions add per-node retry with exponential backoff, 35+ new native integrations (including dedicated Anthropic Claude, Google Gemini, Groq, Perplexity and ElevenLabs nodes for a total of 500+), expanded admin controls, public API access and more enterprise auth options. Cloud plans now bundle monthly AI credits: 2,300 on Starter, up to 13,700 on Pro.
Pricing is genuinely tiered. The self-hosted Community Edition is free ($0) with unlimited executions and every integration, plus a 14-day cloud trial. Paid cloud tiers run Starter at EUR 24/mo, Pro at EUR 60/mo, Business at EUR 800/mo and custom Enterprise; annual billing is roughly 17% cheaper (about EUR 20, 50 and 667 respectively). A start-up program offers 50% off Business for companies under 20 employees.
The main limitations are honest ones: n8n is fair-code under the Sustainable Use License, not OSI-approved open source, so commercial resale is restricted. Self-hosting means you own the maintenance, and the paid cloud pricing steps up sharply between Pro and Business.
It suits developers and technical teams who want automation they can host and audit themselves, and who increasingly want to build AI-agent workflows without stitching together separate tools.
- Free self-hosted Community Edition with unlimited executions and all 500+ integrations
- Rebuilt AI Agent node supports Claude, GPT-4o, Gemini, Mistral and Groq with four memory options
- New spatial Canvas UI and per-node retry with exponential backoff make complex flows manageable
- You can self-host for full data control or use managed cloud with bundled AI credits
- Visual editor plus optional code (JavaScript/Python) fits both non-coders and developers
- Fair-code under the Sustainable Use License, not true OSI open source, so resale is restricted
- Cloud pricing jumps steeply from Pro (EUR 60) to Business (EUR 800)
- Self-hosting shifts maintenance, updates and scaling onto you
- AI credits on cloud plans are capped and can run out on heavy agent workloads
- Steeper learning curve than pure no-code tools once workflows get complex
Choose n8n if…
- You want to self-host automation for full data control at zero license cost
- You are building AI-agent workflows with Claude, GPT-4o or Gemini and need memory options
- You are a developer who wants visual flows plus the escape hatch of real code
- You need 500+ integrations without per-execution pricing on the free tier
Leave a Review
Reviews are published after moderation. We don't share your email.