2026 Review

n8n

Source-available workflow automation with native AI agent nodes — self-host free or run in the cloud.

AutomationFreemiumAPIPopular85% positiveAgent

Facts checked Aug 2, 2026

About n8n

n8n is a source-available workflow automation platform loved by technical teams: build in a visual editor, drop into JavaScript or Python when you need to, and chain LLM calls, tools, and memory into full AI agents with native nodes. The Community Edition is free to self-host with no execution caps; n8n Cloud starts at €24/month (€20/month billed annually) for 2,500 executions, with a free Pro-feature trial and no card required.

At a glance

Founded: 2019
Headquarters: Berlin, DE
Users: 3,000+ enterprise customers (company claim)
Platforms: Web, CLI

Integrates with

OpenAIAnthropicSlackGoogle SheetsPostgresTelegram

Main competitors

Zapier MakeFlowise

Best for

AI agentsAPI orchestrationInternal toolsData pipelines

Pros

Self-hosting is genuinely free — unlimited workflows and executions on your own server
Best-in-class AI agent building among automation tools, with swappable LLMs
Executions pricing beats per-task billing for complex workflows — one execution can run hundreds of steps
Around 200,000 GitHub stars and a large template library mean answers are easy to find

Cons

Steepest learning curve of the big three automation tools — expect to think like a developer
Cloud Starter's 2,500 executions/month evaporates fast with polling triggers, and workflows can pause when you hit the cap
Far fewer native app integrations than Zapier's 9,000+
Self-hosting shifts uptime, updates, and security onto you

Our verdict

Best for: Developers and technical teams building AI agents or complex integrations who want control — and possibly their own infrastructure.

Skip if: Nobody on the team writes code — Make or Zapier will get you shipping automations much faster.

Our review

n8n has become the default answer to "how do I build an AI agent without writing a whole backend." Its AI Agent nodes — with tools, memory, and any LLM from OpenAI to Anthropic to a local model — are genuinely first-class, not a chatbot bolted onto an automation tool. Add execution-based pricing, where a 200-step workflow costs one execution, and complex builds get dramatically cheaper than on Zapier or Make. The kicker is self-hosting: the Community Edition is free, unlimited, and keeps data on your infrastructure.

The cost is approachability. n8n assumes you're comfortable with JSON, expressions, and the occasional code node; non-technical teammates will stall where they'd cruise in Zapier. Cloud quotas also bite — 2,500 Starter executions sound generous until a 5-minute polling trigger eats most of them, and workflows can pause mid-month when you hit the cap.

The company's momentum is hard to ignore — a $180M Series C in late 2025, roughly 200,000 GitHub stars, and a template ecosystem that has made it the automation layer of choice for the AI-agent crowd.

Verdict: the best tool in this directory for AI agent workflows and the only serious one you can run entirely on your own servers. If nobody on the team codes, pick Make or Zapier instead.

What reviewers elsewhere say

85% positive — Technical users rave about flexibility, AI agent nodes, and self-hosting; the recurring gripes are the learning curve and cloud execution caps.

G2 4.8 out of 5 (238 reviews)Capterra 4.6 out of 5

Sentiment estimated from reviews on G2, Reddit, Product Hunt, as of Aug 2026. Third-party scores are point-in-time snapshots. We don't host user reviews.

Key features

AI Agent nodes with tools, memory, and any LLM — OpenAI, Anthropic, or local models
Visual editor with inline JavaScript and Python code steps
Free self-hosted Community Edition with unlimited workflows
500+ integrations plus generic HTTP, webhook, and MCP support
Execution-based cloud pricing — a 50-step workflow costs the same as a 2-step one
Version control, environments, and SSO on higher tiers

n8n as an agent

Agent platformRuns unattendedSomewhat predictable cost

Job: Agent builders
Runs on: Self-hosted, Vendor cloud
Your own model key: Supported

What you give it access to

Your files
Email and calendar
Code repositories

What to watch for

You build the agent, so its guardrails are yours to get right
Self-hosting means you own security, updates and uptime
Model tokens are billed by your provider on top of the platform

How it bills

Usage (tokens/minutes). One task here means one workflow execution, however many nodes and agent steps it contains. Cloud plans bill per full workflow execution rather than per step, so a forty-node run costs the same as a two-node one; self-hosting moves the cost to your own infrastructure plus model tokens.

Agent capabilities checked August 2026. Not hands-on tested — these are the vendor's documented capabilities, read carefully. Limits and permissions change between checks, so confirm approval settings and any spending cap before you connect an account. Compare every AI agent.

How to use n8n

Start a free cloud trial (no card) or self-host the Community Edition via Docker or npm.
Create a workflow and add a trigger — webhook, schedule, or app event.
Chain nodes visually, dropping into JavaScript or Python where logic gets hairy.
For agents, add an AI Agent node, connect an LLM, and give it tools and memory.
Test with pinned data, then activate and monitor executions.
Move to annual billing or self-hosting once your execution volume stabilizes.

Pricing

Deal: Community Edition is free to self-host with unlimited workflows and executions

Community$0

Free self-hosted, source-available. Unlimited workflows and executions on your own infrastructure.

Starter€24/mo

€20/mo billed annually. 2,500 executions/month, 5 concurrent, unlimited workflows and users.

Pro€60/mo

€50/mo billed annually. 10,000 executions/month, 20 concurrent, admin roles, insights.

Business€667/mo

40,000 executions/month, SSO/SAML, environments, self-hosted option.

EnterpriseCustom

Custom executions, hosted by n8n or self-hosted, dedicated support.

Pricing verified August 2026 · prices as listed by the vendor; check the official site for current offers.

Try n8n

Frequently asked questions

What is n8n?

n8n is a source-available workflow automation platform with a visual editor, inline code steps, and native AI agent nodes. You can self-host it free or use n8n Cloud.

How much does n8n cost?

The self-hosted Community Edition is free with unlimited executions. Cloud starts at €24/month (€20/month billed annually) for 2,500 executions; Pro is €60/month (€50 annually) for 10,000.

Is n8n really free?

Yes, if you self-host — the Community Edition has no execution or workflow limits. You pay only for your own server and handle updates and security yourself.

Does n8n have an API?

Yes — n8n's public REST API lets you manage workflows, executions, and credentials programmatically, and n8n can act as both an MCP client and server for AI assistants.

n8n vs Zapier or Make?

n8n wins on AI agents, code-level control, price per complex workflow, and self-hosting. Zapier wins on app coverage (9,000+ apps) and ease; Make sits between them with its visual canvas.

n8n