Dagu is a self-contained, single-binary workflow engine for defining DAGs in YAML. Free and open source, with optional paid enterprise licenses.
Researched and fact-checked by Bartu Cavusoglu · Updated July 18, 2026
Dagu is a lightweight, self-contained workflow orchestration engine, written in Go, that runs as a single binary with no external database requirement, positioned explicitly as an alternative to Apache Airflow and cron for teams that don't want to run an SRE-heavy orchestration platform.
Workflows are defined as YAML-based DAGs supporting shell, Docker, Kubernetes, SSH, HTTP, and SQL steps, with a built-in web UI, scheduling, retries, and — increasingly — a Model Context Protocol server so AI coding agents can author and run workflows directly.
Dagu supports three deployment modes — standalone, headless, and distributed with a coordinator and labeled workers — letting it scale from a single laptop to a small fleet without switching products.
It is used in production for ETL, legacy script orchestration, media conversion, infrastructure automation, containerized batch jobs, and AI-agent-driven task execution.
The community edition is GPLv3-licensed and free forever, supporting unlimited self-hosted servers and workers with core scheduling and execution features.
Paid self-hosted licenses add enterprise controls: Minimum is $50/month (up to 3 servers), Pro is $150/month (up to 15 servers), and Enterprise is custom-priced; a managed GKE-hosted cloud option is also available.
Yes, Dagu's community edition is GPLv3-licensed and free forever, with unlimited self-hosted servers and workers and core scheduling/execution features included.
No. Dagu stores execution state in local files by default and runs as a single binary with no required external database or message broker.
Yes, Dagu has native executors for Docker containers, Kubernetes jobs, SSH commands, HTTP requests, and SQL queries in addition to shell commands.
Paid self-hosted licenses (Minimum at $50/month and Pro at $150/month) add enterprise features like SSO, RBAC, and audit logging on top of the free core engine.
Yes, Dagu includes a built-in Model Context Protocol (MCP) server that lets AI coding agents such as Claude Code and Gemini CLI author and run workflows directly.
Dagu is designed to avoid Airflow's operational complexity, running as a single dependency-free binary instead of requiring a database, broker, and multiple long-running services.