Plain-AyI
← All projects
Training & fine-tuning

Agent Lightning

A Microsoft framework for training AI agents to get better at tasks through trial and reward.

Research-backed

What it does

Agent Lightning uses reinforcement learning, where an AI improves by trying tasks and being rewarded for good results. It trains agents while they run in their normal setup, without changing their code.

Its makers report that this raised a coding model's score on a programming benchmark after a few thousand training examples. They describe the core as small, about 3,500 lines of code.

Who it's for

AI researchers and engineers who train agents.

Can I use it without coding?

Developers only. Needs powerful GPUs and machine-learning expertise. New to this? Start here.

Can I run it?

Cost
FreeNothing to pay or sign up for.
Computer
Works best on a powerful computer or one with a graphics card (GPU)
API key
Not needed

Try it

Using Claude Code, Cursor or another AI coding assistant? Paste the prompt and it will walk you through the setup.New to this?

Before you install

  • Only download from the official GitHub page. Copies with similar names exist for many popular projects.
  • License (MIT): Very permissive: free to use, change and share, including for business.
  • More on staying safe

Where people are talking about it