Plain-AyI
← All projects
Robotics & simulation

Unity ML-Agents Toolkit

Turns Unity games and 3D simulations into practice worlds where AI characters learn by trial and error.

Research-backed

What it does

ML-Agents lets game developers, researchers and hobbyists train AI "agents", characters or robots that learn behaviours, inside games and simulations built with the Unity game engine.

Agents can learn by trial and error with rewards (reinforcement learning) or by copying recorded demonstrations (imitation learning). Several agents can learn to cooperate or compete, and tasks can start easy and get harder over time. It comes with many example environments and uses PyTorch behind the scenes.

Trained behaviours can then run inside Unity games on many platforms. The README notes that Unity collects some usage data inside the editor.

Who it's for

Game developers, AI researchers and students exploring how agents learn.

Can I use it without coding?

Developers only. Needs the Unity editor, Python and some coding. New to this? Start here.

Can I run it?

Cost
FreeNothing to pay or sign up for.
API key
Not needed

Try it

Official download page

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 or the official download page above. Copies with similar names exist for many popular projects.
  • No standard license listed. Check the GitHub page before using it for business.
  • More on staying safe

Where people are talking about it