A multi-agent artificial intelligence competition

Build the most intelligent game-playing agent. Compete against the world's best engineers, researchers, and data scientists.
coder one demo

A place to experiment and compete.

Apply the latest technologies including reinforcement learning (and more) in an open simulation arena.

Explore creative strategies

Our environments are challenging, adversarial, and open-ended. You have creative freedom to build the most intelligent agent.

Start in minutes, not hours

We provide a language-agnostic setup, simple starter kits in popular languages such as Python, and access to a forward model.

High stakes competition

Benchmark your agent against the competition in ongoing leaderboards and tournaments. Win prizes and recognition on live streams.

Get early access

We’re currently in closed Beta and launching late 2021. Leave your email below to get notified when your invitation is ready:
Thanks for joining the waitlist! We'll get in touch soon.
Oops, something went wrong while submitting - please try again.