arcade-learning-environment
« Back to VersTracker
Description:
Platform for AI research
Type: Formula  |  Latest Version: 0.11.2@2  |  Tracked Since: Dec 17, 2025
Links: Homepage  |  formulae.brew.sh
Category: Ai ml
Tags: reinforcement-learning ai atari research gym
Install: brew install arcade-learning-environment
About:
The Arcade Learning Environment (ALE) is a Python framework that provides a high-level interface to Atari 2600 games, serving as a standard benchmark environment for reinforcement learning agents. It isolates agents from the emulator details, allowing researchers to focus on algorithm development. Its main value is providing a challenging, reproducible testbed for evaluating general AI performance.
Key Features:
  • Provides a standardized API for reinforcement learning agents
  • Supports frame stacking and observation preprocessing
  • Includes a diverse set of Atari 2600 game ROMs
  • Offers stochasticity via random frame skips for robust evaluation
  • Hardware-accelerated rendering via SDL2 and OpenGL
Use Cases:
  • Benchmarking deep reinforcement learning algorithms
  • Researching general value functions and agent capabilities
  • Educational projects in AI and machine learning
  • Developing agents that learn from pixel input
Alternatives:
  • Gymnasium – ALE is often used as the backend environment within Gymnasium, which provides a broader collection of RL environments and a unified API.
  • PyGame Learning Environment – A simpler 2D grid-world environment, whereas ALE provides the complexity and visual richness of classic video games.
License: GPL-2.0-only
Dependencies: numpy, opencv, python@3.14, sdl2
Bottles available for: arm64_tahoe, arm64_sequoia, arm64_sonoma, sonoma, arm64_linux, x86_64_linux
Version History
Detected Version Rev Change Commit
Dec 13, 2025 5:36pm 2 VERSION_BUMP 211ec3eb
Oct 29, 2025 3:59am 1 VERSION_BUMP beee7275
Sep 18, 2025 10:33am 1 VERSION_BUMP 9b52bdb6
Sep 16, 2025 10:14pm 1 VERSION_BUMP 61a9a20a
Sep 28, 2024 10:30am 0 VERSION_BUMP 6b366214