Platform · ai agent automation
n8n
Source-available workflow automation you can self-host, with native LLM nodes
The verdict
n8n is a workflow automation tool you can run on your own server, building flows on a visual canvas. It ships LLM and agent nodes, so AI steps sit alongside ordinary app actions.
What it is
n8n builds automations as nodes on a canvas, where each node is an app action, a function, or an AI step. Because you can self-host it, high-volume workflows avoid per-task metering and sensitive data can stay in your own infrastructure. Native nodes for OpenAI, Anthropic and LangChain let you add an agent or an LLM call mid-flow. The catch is operational: you maintain the server, updates and reliability yourself unless you pay for the cloud version, and the connector count trails Zapier's.
Strengths
- Self-hosting removes per-task fees, so high-frequency flows stay cheap at scale
- Native OpenAI, Anthropic and LangChain nodes build AI agents inside a normal workflow
- An HTTP node and function steps let you call any API the connector library lacks
Trade-offs
- Self-hosting means you own uptime, updates and backups, which is real ongoing work
- The connector library is smaller than Zapier's, so some niche apps need manual HTTP setup
- The canvas has a steeper first-hour learning curve than a simple trigger-action builder
Best for
High-volume or data-sensitive automations where self-hosting cuts cost and keeps data in-house
Not for
Non-technical users who want a managed service and the broadest possible app library with zero maintenance
AI tools that fit n8n
Scored on how well each tool actually works with n8n — integration depth, workflow fit, and real adoption. Higher is better.
The AI ecosystem
Around 1,000+ nodes plus a generic HTTP node, and first-party AI/LangChain nodes for agent building.
Pricing
Self-hosted (free, open-source) plus n8n Cloud paid plans (Starter, Pro, Enterprise) priced by executions/features, roughly from the low-to-mid tens of dollars per month upward.
Community (self-host)
Free$0
- self-hosted
- unlimited executions on your hardware
Cloud Starter
$24/mo
- managed hosting
- execution quota
Cloud Pro
$60/mo
- more executions
- advanced features
Enterprise
Custom
- SSO
- scaling
- support
What people build with it
Frequently asked
- Can n8n run AI agents?
- Yes. n8n has native AI Agent and LangChain nodes plus connectors for OpenAI and Anthropic, so agents and LLM steps live inside ordinary workflows.
- Is n8n free?
- The self-hosted Community edition is free to run on your own server; managed cloud plans are paid and start in the low tens of dollars per month.
- How is n8n different from Zapier?
- n8n can be self-hosted to avoid per-task fees and keep data in-house, but it has fewer ready-made connectors and requires more technical setup.
- Does n8n require coding?
- Basic flows are built visually, but advanced cases use function and HTTP nodes, so some technical comfort helps more than with simpler tools.
- Can n8n connect to apps it doesn't list?
- Yes. The generic HTTP Request node calls any REST API, so unlisted services can still be automated with manual configuration.