Skip to content

Developer tools

Airflow

Discover Apache Airflow, the ultimate platform for programmatically authoring, scheduling, and monitoring workflows. Transform complex tasks into manageable, streamlined operations with dynamic and extensible DAGs. Enhance your workflow efficiency today!

Deploy Airflow on Open Source Cloud

  1. Web console: press Launch above, sign in, name the instance and create it.
  2. Command line: with your personal access token in OSC_ACCESS_TOKEN, run npx -y @osaas/cli@latest create apache-airflow myinstance. Add -o Key=Value for any required settings.
  3. Your AI agent: connect it to the OSC MCP server at https://mcp.osaas.io/mcp and ask: β€œCreate an instance of Airflow on Open Source Cloud and give me its URL.” How to connect

What it costs

One running Airflow instance uses 10 tokens a day. The free Basic plan's 100 one-time tokens run it for about 10 days. On Personal (40 tokens a day) one instance runs continuously.

See plans and token allowances

Open source, no lock-in

Airflow is built from the open source project at github.com/apache/airflow. You can run the same project on your own infrastructure or another cloud whenever you want, and take your data with you.

Questions about Airflow

How much does Airflow cost on Open Source Cloud?

One running Airflow instance uses 10 tokens a day. The free Basic plan's 100 one-time tokens run it for about 10 days. On Personal (40 tokens a day) one instance runs continuously. Prices and allowances are on https://www.osaas.io/pricing.

How do I deploy Airflow?

Three ways: press Launch to create it in the web console; run npx -y @osaas/cli@latest create apache-airflow myinstance with OSC_ACCESS_TOKEN set (add -o Key=Value for any required settings); or connect your AI agent to the OSC MCP server at https://mcp.osaas.io/mcp and ask it to "Create an instance of Airflow on Open Source Cloud and give me its URL."

Can I move Airflow off Open Source Cloud later?

Yes. Airflow is open source, built from https://github.com/apache/airflow, so you can run the same project on your own infrastructure or another cloud and take your data with you.