The Home of RL Environments

Featured environments
Syncing

Getting started

Start sampling in 2 minutes

1

Sign up

Create a free OpenReward account to use and create environments.

2

Get your API key

Generate an API key to make environment requests.

3

Choose a model provider

Use public or your own models to sample from the environment.

4

Sample from environments

Call environment endpoints and start making rollouts!

Try it live

Sample GeneralReasoning/MLEBench-Train

Our Team

Experience at scale

Previously our team built state-of-the-art models, evaluations and products used by hundreds of millions of people.

LLMs

  • Llama 2
  • Llama 3
  • Galactica
  • GPT-4o

Benchmarks

  • MLE-Bench
  • PaperBench
  • SWE-Bench Verified

Products & platforms

  • Papers with Code

Backed by the best

Legends who built the tools we love

Soumith Chintala

Creator of PyTorch

Jeff Dean

Chief Scientist, Google DeepMind

Joe Spisak

Product Director, Meta Superintelligence Labs

Laurens van der Maaten

Member of Technical Staff, Anthropic

Robert Stojnic

Co-creator of Papers with Code

Logan Kilpatrick

Product lead, Google DeepMind