n8n
Automation · Freemium
Overview
A workflow automation tool that you can run on your own server. You build flows by connecting nodes on a canvas, where each node reads from or acts on a service. It was started in 2019 by Jan Oberhauser in Germany, and the company is now n8n GmbH. The code is public on GitHub under a source-available license called the Sustainable Use License, which is not a standard open-source license. Alongside the free self-hosted Community Edition, n8n sells a hosted cloud version and paid self-hosted tiers. It has become popular for building AI agent workflows.
What people use it for
Technical teams self-host n8n so that regulated or sensitive data stays on their own infrastructure and never passes through a third-party automation vendor. They use it to sync data between internal systems, databases, and SaaS tools, and to run scheduled jobs such as reports and cleanups. It is widely used to build AI workflows: chaining model calls, adding retrieval from a vector store, routing between agents, and pausing for a human to approve a step. Developers use the code node to drop in JavaScript or Python where a visual node is not enough. Agencies and consultants build automations for clients on their own hosted instance. Some teams pick it to avoid per-task pricing on high-volume automations, since cloud plans bill by workflow execution rather than by step.
Key capabilities
n8n ships more than 500 integration nodes and can call any other service over HTTP. The code node runs JavaScript or Python for custom logic, and workflows can be triggered on a schedule, by webhook, or by an event in another system. For AI work it has nodes for LangChain-style agents, retrieval-augmented generation, multiple model providers, and human-in-the-loop steps, with the ability to swap between cloud and local models. The full source is on GitHub, with more than 200,000 stars, and can be self-hosted with Docker, including air-gapped setups. The Community Edition is free to run under the Sustainable Use License. Cloud and paid self-hosted plans bill by full workflow execution, not per node, and include unlimited users, workflows, and integrations. Paid tiers add SSO, Git-based version control of workflows, and separate environments; Enterprise adds external secret stores and log streaming.
Limitations
Self-hosting means you own the upkeep: server setup, upgrades, backups, scaling, and fixing it when a workflow queue stalls at night. The hosted cloud version removes that work but costs more and undoes the main reason many people chose n8n. The Sustainable Use License restricts reselling n8n or hosting it as a service for others, so it is not as free as "open source" suggests. The interface is less polished than some paid competitors and error messages can be cryptic. Some community-contributed nodes are unmaintained. Documentation is broad but thin in places, so hard problems often end up in the community forum.
Insight
The one clear reason to choose n8n is control. If data cannot leave your infrastructure for legal or contractual reasons, n8n on your own server solves that in a way a cloud tool cannot. It has also become a common home for AI agent workflows, partly because the nodes are kept current and partly because the community publishes a lot of templates. The trade-off is real. Self-hosting is a standing commitment to patching, monitoring, and recovery, and the moment you move to their cloud to avoid that, a cheaper cloud competitor may do the job. It also is not truly open source, so read the license before building a business on it. Best for teams with in-house technical skill and a concrete need to keep data close.
Pricing
Freemium
Website
Last checked 2026-08-30