Farama-Foundation/stable-retro
Retro games for Reinforcement Learning Research
This project helps reinforcement learning researchers convert classic video games into standardized environments for training AI agents. You provide ROM files for games like Super Mario World or Street Fighter II, and the project outputs a structured interface that your AI models can interact with, observing game states and executing actions. It's designed for machine learning researchers and students working on reinforcement learning algorithms.
354 stars.
Use this if you are a reinforcement learning researcher looking to train AI models on a wide array of classic video games.
Not ideal if you are a casual gamer looking for an emulator to simply play retro games.
Stars
354
Forks
67
Language
C++
License
MIT
Category
Last pushed
Mar 09, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/Farama-Foundation/stable-retro"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
MatPoliquin/stable-retro-scripts
Train models on retro games. AI vs AI contest. Pytorch C++ plugin for RetroArch that let you...
nadavbh12/Retro-Learning-Environment
The Retro Learning Environment (RLE) -- a learning framework for AI
svpino/cs7641-assignment4
CS7641 - Machine Learning - Assignment 4 - Markov Decision Processes
Avalon-Benchmark/avalon
A 3D video game environment and benchmark designed from scratch for reinforcement learning research
abhinavcreed13/ai-reinforcement-learning
This project will implement value iteration and Q-learning. It will first test agents on...