n8n Review (2026): Features, Pricing & Alternatives

What is n8n?

n8n is a node-based automation tool that connects apps, webhooks, and AI models. The important pricing fact is what you are billed for: a full workflow run ('execution'), not each node and not each seat.

Community Edition on GitHub is free to self-host. n8n Cloud, as of September 2026, lists Starter at €20/month billed annually (about €24 month-to-month) with 2,500 executions, 5 concurrent runs, one shared project, and 2,300 AI Assistant credits.

Pro is €50/month billed annually (about €60 month-to-month); help-center tiers split Pro-1 (10,000 executions, 20 concurrent, 5,700 AI credits) and Pro-2 (50,000 executions, 50 concurrent, 13,700 AI credits). Extra executions on monthly Pro-2 are about €2 per 1,000, up to 500,000.

Business is €667/month billed annually and is currently self-hosted only (SSO, Git version control, environments). Enterprise is quote-only.

Cloud data is hosted in Frankfurt. Unused AI credits do not roll over.

A concrete walkthrough: webhook lead → CRM → Slack without burning 2,500 runs

One webhook that fires on every form submit is one execution, even if the workflow has twelve nodes (enrich, filter, HubSpot, Slack). That is the whole point of n8n's meter.

A form that gets 40 leads a day is about 1,200 executions/month — Cloud Starter (€20 yearly / ~€24 monthly) still has headroom. A workflow that runs every 5 minutes is about 8,600 executions/month and will blow Starter; that job belongs on Pro-1 or on self-hosted Community.

Do not put the AI Assistant on the critical path — those Cloud credits reset on the 1st and do not roll over. If the payload cannot leave your VPC, skip Cloud and run Community or Business on your own box (Cloud data sits in Frankfurt).

Key features

Visual node editor with 400+ integrations, plus JS/Python code nodes when a connector is missing
Billed per full workflow execution, not per step or per user — unlimited users on every Cloud plan
Native AI nodes and an AI Assistant (preview) with a monthly credit pot on Cloud
Self-host Community for free, or Cloud if you do not want to run Docker yourself
Error handling, retries, schedules, webhooks, and JSON export that diffs in git
Business/Enterprise add SSO, Git version control, separate environments, and (Enterprise) log streaming

How to get started

If you can run Docker, start with Community and keep data on your own server — that is genuinely $0
Otherwise start an n8n Cloud trial (1,000 executions, Pro-like editor features, no card on the Cloud trial)
Add a trigger (schedule, webhook, or app event), then nodes for transform / HTTP / AI
Test with sample data, then activate; count production runs in Insights, not in your head
Upgrade Starter → Pro when 2,500 executions or 5 concurrent runs are the wall; do not upgrade for 'more users'

n8n pricing

From n8n.io/pricing and the Cloud FAQ, September 2026. Community self-host is free.

Cloud Starter is €20/month billed annually (~€24 monthly), 2,500 executions, 2,300 AI credits. Cloud Pro is €50/month billed annually (~€60 monthly); Pro-1 is 10k executions, Pro-2 is 50k.

Business is €667/month billed annually, self-hosted at the time of review. Enterprise is custom.

Annual billing changes price, not the execution cap.

Plan	Price	Best for
Community (self-host)	Free	Full editor on your own Docker/VPS
Cloud Starter	€20/mo billed yearly (~€24 monthly)	2,500 executions, 5 concurrent, managed host
Cloud Pro-1	€50/mo billed yearly (~€60 monthly)	10k executions, admin roles, workflow history
Cloud Pro-2	Same Pro family, higher usage card	50k executions, 50 concurrent; extra €2 / 1k runs
Business	€667/mo billed yearly	Self-host + SSO, Git, environments
Enterprise	Custom	SLA, log streaming, 200+ concurrent

Cloud bills complete workflow runs, not nodes. AI Assistant credits reset on the 1st and do not roll over.

Our take: Self-host if you have ops time — Community is the real free plan. Use Cloud Starter to avoid servers until 2,500 monthly runs. Buy Pro for concurrency, history, and admin, not because you hired a second person (users are already unlimited).

Prices verified 2026-09. Plans change often — confirm on the official site above before you buy.

Use cases

Lead intake: form webhook → enrich → write CRM → Slack ping, as one execution
An AI agent that calls several APIs inside one workflow instead of a custom backend
Scheduled reports that pull three tools and email a summary
Internal tools that need a code node Zapier would not let you write

Strengths and tradeoffs

Where it stands out

One execution = one full run, so a 15-node workflow costs the same as a 2-node workflow
Users are unlimited on every Cloud plan; you do not pay Zapier-style per seat
Community self-host is actually free, not a demo with a badge

Tradeoffs

Cloud Starter's 2,500 executions vanish fast on tight cron jobs (every 5 minutes ≈ 8.6k/month)
Business was Cloud-unavailable at review time — SSO/Git meant self-hosting
AI Assistant credits do not roll over and cannot be bought a la carte

n8n alternatives

Zapier AI: Zapier is easier for non-technical teams and bills per task, which gets expensive on chatty multi-step zaps.
Make: Make has a more polished visual builder and is still a hosted SaaS — no first-class free self-host.
Dify: Dify is an AI-app builder (RAG, chat widgets) rather than a general automation bus.
LangGraph: LangGraph is a code-first agent graph you deploy yourself; n8n is the visual ops layer.

Pick n8n when you want a self-host option or execution-based pricing. Pick Zapier when the team will not touch nodes or Docker.

Pick Make for a prettier hosted canvas. Pick Dify or LangGraph when the artifact is an AI app or an agent, not a business workflow.

FAQ

Is n8n free to use?

n8n's plans: Free Community self-host; Cloud from €20/mo billed annually. Pricing and free-tier limits change over time, so check the official site above for the latest details.

What is n8n used for?

Open-source workflow automation you can self-host, or run on n8n Cloud by execution count.

Who is n8n best for?

n8n is a good fit for lead intake: form webhook → enrich → write CRM → Slack ping, as one execution, or an AI agent that calls several APIs inside one workflow instead of a custom backend.

What are the alternatives to n8n?

Commonly compared alternatives include Zapier AI, Make, and Dify, see the comparison above for how they differ.

How much does n8n cost?

Self-hosting Community is free. n8n Cloud starts at €20/month billed annually for Starter (2,500 executions). Pro is €50/month billed annually. Business is €667/month billed annually and was self-hosted-only when we checked. You pay per workflow run, not per seat.

What's the best n8n alternative?

Zapier AI is the most commonly recommended swap: zapier is easier for non-technical teams and bills per task, which gets expensive on chatty multi-step zaps. Which one is actually best depends on which of those tradeoffs matters more for what you're doing.

Is n8n really free?

Community Edition is free to self-host. n8n Cloud is paid and starts at €20/month billed annually for Starter. Business features (SSO, Git environments) needed a self-hosted license when we reviewed the pricing page.

Do I pay per user or per node?

Neither. Cloud plans include unlimited users and you pay for a complete workflow execution. Annual vs monthly only changes the euro price, not the execution cap.

n8n