Flyte OSS
Flyte 2 is the durable AI runtime, in open source.
You write training, serving, and agent workloads in plain Python, and Flyte can recover them from failure, whether your code raised an error or the infrastructure failed under it with an OOM kill or a preempted node.
Flyte is Apache 2.0 licensed and a Linux Foundation AI & Data project, and you run it yourself on your own Kubernetes cluster.
Try Flyte on Union.ai, free for 30 days
Union.ai is the commercial platform built on Flyte.
Subscribe to Union Team on AWS Marketplace and try it free for 30 days.
See Setup from Marketplace.
Learn the basics of Flyte, covering all the core concepts around tasks, apps, agents, and typed model calls.
Get started
What Flyte 2 is, how to install it, the core concepts, and the ways to run your code.
Tasks
Configure, build, and deploy the durable batch workloads that everything else is made of.
Apps
Long-running services for dashboards, REST APIs, and model endpoints.
Triggers
Run tasks automatically on a schedule or in reaction to new data, or save named launch configurations to fire on demand.
Agents
Durable, self-healing agents built from tasks and apps, with sandboxing and MCP.
System one types
Answer many typed questions in one call, with calibrated confidence, and decide what happens next in code.
Organize your codebase, optimize performance for production, and migrate from
other workflow orchestrators.
Project patterns
Patterns for BYO images, monorepos with uv, CI/CD, and multi-team resource management.
Software development lifecycle
Automate your own development process with Flyte, using event-driven runs, review and release gates, human approvals, and evaluation gates for systems whose behavior depends on data and models.
Run scaling
Tune task overhead, batching, reusable containers, and fanout to scale your workflows.
Advanced project
An advanced guide for building an LLM reporting agent on Flyte.
Migration
Port a Flyte 1 codebase to Flyte 2, or map Airflow concepts to their Flyte 2 equivalents.
