n8n review: fair-code workflow automation for technical teams

What n8n does and how it differs

n8n is a workflow automation platform that connects applications, APIs, and services through a visual node-based editor. Unlike cloud-only competitors such as Zapier or Make, n8n offers both a managed cloud service and a self-hostable version under a fair-code license, allowing organizations to run unlimited workflows on their own infrastructure without per-execution fees. The platform supports over 400 pre-built integration nodes and allows connection to any API through its HTTP Request node or custom code written in JavaScript or Python.

The fair-code licensing model sits between traditional open source and proprietary SaaS. Under the Sustainable Use License, users can inspect, modify, and self-host n8n for internal business purposes without restriction, but cannot resell it as a competing hosted service. This approach prevents exploitation by large cloud providers while keeping the software accessible to actual users. Enterprise features marked with .ee in filenames require a separate commercial license.

n8n counts usage by workflow executions rather than individual steps or tasks. A single workflow run counts as one execution whether it contains three nodes or thirty, making complex multi-step automations dramatically cheaper than task-based competitors. For example, a 10-step workflow running 10,000 times per month consumes 10,000 executions on n8n but 100,000 billable tasks on Zapier, resulting in cost differences that widen as workflow complexity increases.

Self-hosting versus cloud: the real trade-off

Self-hosted n8n Community Edition is free to run indefinitely with no execution limits, but infrastructure costs and maintenance time are not. A minimal production deployment requires 2–4 vCPU, 4–8GB RAM, and at least 20GB SSD storage, translating to roughly $5–20 per month on VPS providers like Hetzner, DigitalOcean, or AWS. This raw cost underestimates the full picture: SSL certificate management, PostgreSQL database setup and backups, Docker configuration, reverse proxy setup, and ongoing security updates require either in-house DevOps skills or external consulting hours.

Reports from production users indicate initial setup taking anywhere from a few hours for experienced Docker users to multiple days for teams new to containerization. Ongoing maintenance—applying updates, pruning execution logs, monitoring resource usage—adds roughly half a day per month once monitoring is configured. The question is not whether self-hosting is cheaper on paper, but whether the team has spare DevOps capacity or values infrastructure control enough to justify the time investment.

n8n Cloud removes all infrastructure burden at the cost of execution-based metering. As of mid-2026, the Starter plan provides 2,500 monthly executions for €20/month, Pro offers 10,000 for €50/month, and the new Business tier delivers 40,000 for €667/month annually. For teams running fewer than a few thousand executions monthly or lacking server management expertise, Cloud pricing remains competitive with self-hosting once labor costs are factored in. For high-volume users running tens of thousands of executions, self-hosting delivers order-of-magnitude cost savings despite the operational overhead.

Building workflows: visual editor meets code

n8n's workflow canvas uses a node-based interface where each block represents a trigger, action, or transformation. The editor displays real-time execution previews, showing output data next to each node as the workflow runs, which significantly speeds debugging compared to linear tools like Zapier. Users can execute individual nodes in isolation without running the entire workflow, a feature praised in reviews for reducing iteration time when building complex logic.

The platform expects some technical literacy. While drag-and-drop node placement is straightforward, configuring nodes often requires understanding JSON data structures, HTTP methods, and authentication protocols. The documentation provides detailed node references and API guides but assumes familiarity with concepts like OAuth2 flows and webhook payloads. Non-technical users report a steeper learning curve than Zapier's guided setup, though many note that AI coding assistants now help bridge the gap for JavaScript expression writing.

Code nodes allow embedding custom JavaScript or Python directly into workflows, removing the need to build external scripts or wait for official integration updates. The HTTP Request node converts cURL commands from API documentation into working nodes automatically, a capability that effectively extends n8n's integration library to any REST API. Reviews consistently highlight this flexibility as n8n's core differentiator, enabling connections to niche European SaaS tools, internal company APIs, and legacy systems without pre-built connectors.

AI workflows and LangChain integration

n8n includes native AI capabilities through dedicated nodes for major LLM providers (OpenAI, Anthropic, Google), vector databases, and full LangChain integration for building multi-step agent workflows. Users can construct retrieval-augmented generation pipelines, multi-agent systems with tool access, and chatbot interfaces entirely within the visual editor. The platform supports both cloud-based models and self-hosted LLMs, giving teams control over which provider handles sensitive data.

The AI Assistant feature, currently in early preview on n8n Cloud, generates workflow suggestions from natural language descriptions and offers iterative refinement through chat. Monthly AI credit allowances (2,300 on Starter, up to 13,700 on higher Pro tiers) limit usage, though self-hosted instances will allow users to bring their own API keys without credit caps once the feature reaches general availability. This positions n8n between Zapier's simpler AI actions and full custom agent frameworks requiring Python expertise.

Reviews from teams building AI-powered automation praise n8n's ability to mix deterministic workflow logic with AI reasoning exactly where needed, avoiding the all-or-nothing approach of fully autonomous agents. The visual canvas makes complex agent orchestration easier to debug and audit than code-only frameworks, a critical advantage for production deployments where explainability and error handling matter. However, advanced AI use cases involving custom embeddings or fine-tuned models still require code node implementation rather than pre-built components.

Community health and vendor trajectory

n8n's GitHub repository shows strong development velocity with over 150,000 stars as of late 2025, growing from 100,000 in mid-2025. The project maintains active daily commits, has surpassed 100 million Docker pulls, and supports a community forum where both n8n team members and experienced users respond to questions, often within hours. A library of over 8,600 community-contributed workflow templates across 25 categories provides ready-made starting points for common automation patterns.

The company behind n8n raised a $55 million Series B led by Highland Capital following Sequoia's initial seed investment, suggesting financial runway for continued development. The fair-code model generates revenue through n8n Cloud subscriptions and Enterprise licenses while keeping the core platform accessible. This structure appears more sustainable than pure open-source projects dependent on donations, though it introduces risk if the company pivots licensing terms—a concern mitigated somewhat by the ability to fork the codebase under current licenses.

Documentation quality receives mixed feedback. Technical references for individual nodes and API integration are comprehensive, but step-by-step tutorials for non-developers remain sparse. The community forum partially fills this gap, with workflow examples and troubleshooting threads, though official support is limited to Enterprise contracts. For self-hosted deployments, GitHub issues serve as the primary channel for bug reports and feature requests, with transparent but sometimes slow responses compared to commercial helpdesk SLAs.