OpenAI released the public beta of OpenAI Gym, its first major release since founding, as a toolkit for developing and comparing reinforcement learning algorithms. It offered a diverse suite of environments, from simulated robots to Atari games, behind a common interface, aimed at the lack of standardized benchmarks in RL research.