Skip to main content
Back

Prefect Acquires Dagster to Build the Ultimate AI Workflow Platform

Prefect is acquiring Dagster Labs, merging the two biggest modern successors to Airflow into one company covering pipelines, MLOps, and agents.

Repo, Full Access
  • Prefect is acquiring Dagster Labs, uniting the two leading modern successors to Apache Airflow. Details
  • Dagster keeps its name, Apache 2.0 open-source license, roadmap, and Dagster+ commercial offering. FAQ
  • Combined company will operate under the Prefect brand after the transaction closes.
  • Nick Schrock and Pete Hunt become strategic advisors to Prefect while staying active in open source.
  • Strategic frame: Dagster defines outcomes, Prefect executes work, FastMCP governs agent access to tools.
  • Pricing for Dagster+ and Prefect Cloud is unchanged; no migration required for existing deployments.

The two loudest rivals in modern workflow orchestration are becoming one company. Prefect has agreed to acquire Dagster Labs, the creator of the asset-based orchestrator that has been the most visible alternative to Prefect itself for close to a decade. The transaction brings together the two most widely adopted successors to Apache Airflow, creating a combined suite that serves thousands of data teams running production workloads across data pipelines, ML operations, and AI agent infrastructure.

If you run Dagster in production, the practical answer to "what changes for me" is: not much, for now. Dagster remains a supported product under its existing name. The open source project continues under its existing license, and Dagster+ remains a supported commercial offering. The combined company plans to operate under the Prefect name, but both open-source projects and both commercial clouds keep shipping as independent products.

Two rivals, one thesis

Prefect and Dagster have spent years converging on the same problem from opposite ends of the workflow spectrum. In the words of Dagster founder Nick Schrock, Prefect is preferred by many machine learning and platform teams. It excels at dynamic workflows, offers an elegant Python API, and has the smoothest onboarding experience in the category. Dagster is the preferred choice for forward-looking data teams, with advanced capabilities in partitioning, lineage, cataloging, and scheduling.

That difference in temperament shows up in the code. Prefect leans imperative: you write Python flows and the runtime executes them. Dagster leans declarative: you define data assets, their dependencies, and the conditions under which they should be materialized, then the platform figures out the schedule.