n8n

n8n (pronounced "nodemation") is the open-source alternative to Zapier and Make, designed for teams that want the power of visual workflow automation combined with the flexibility of code — and the control of self-hosted deployment. With 188,600+ GitHub stars, it has become the go-to automation platform for technical teams at startups and enterprises where data privacy, vendor independence, and custom extensibility matter.

Self-hosted and cloud options

n8n's Community Edition is completely free and self-hosted — you run it on your own server, Docker container, or Kubernetes cluster. All data stays in your environment; there is no SaaS vendor with access to your workflow data. This is particularly important for organizations in regulated industries (healthcare, finance, legal) where third-party data processors must be carefully controlled. The cloud-hosted option (Starter, Pro, Enterprise) provides all the same features without infrastructure management, billed on a per-execution model rather than per-seat or per-task.

• Self-hosted deployment — run on Docker, Kubernetes, or any Linux server; all workflow data stays in your environment.
• Multi-agent AI orchestration — build sophisticated AI agent workflows with multiple LLMs, tool calling, and human-in-the-loop approval steps.
• RAG system support — native integration for retrieval-augmented generation pipelines connecting to vector databases and knowledge bases.
• Visual + code flexibility — switch between visual builder and JavaScript/Python code nodes within the same workflow.
• Human-in-the-loop — pause workflows for manual review and approval before executing irreversible actions.

AI agent workflows

n8n's AI capabilities have grown significantly with the 2024–2025 releases. The platform now supports full multi-agent orchestration: an orchestrator agent can spawn and coordinate sub-agents, each with access to different tools. RAG (retrieval-augmented generation) pipelines can be built visually, connecting n8n to vector databases (Pinecone, Weaviate, Qdrant) and populating LLM context with retrieved knowledge. The LLM evaluation framework allows teams to benchmark agent performance systematically before deploying to production — a capability typically reserved for dedicated MLOps platforms.

Pricing

Community Edition is free forever on self-hosted infrastructure. Cloud Starter at €24/month includes 5,000 executions with email support. Cloud Pro at €60/month covers up to 50,000 executions. Enterprise is custom with unlimited executions and on-premises deployment options. The execution-based pricing (versus per-step/per-task in Zapier and Make) means n8n costs are more predictable for complex workflows where each automation involves many steps — you pay per completed workflow, not per action within the workflow.

Start free at n8n.io or self-host from GitHub.