stable-baselines3 has 1.5M monthly downloads and 14K GitHub stars. Version 2.9.0 supports Python 3.10+, first published in 2020.
It is actively maintained under a permissive license, last released in Jun 2026.
It is built on top of cloudpickle, gymnasium and numpy. It is used by 123 tracked packages, including sb3-contrib, datenwissenschaften and csle-common.
GitHub topics include reinforcement-learning, reinforcement-learning-algorithms and machine-learning.
Package Insights
((week_daily_avg - month_daily_avg) / month_daily_avg) * 100Weekly Downloads
GitHub Stars
Downloads by OS
Python Versions
Top Countries
Dependencies
- cloudpickle
- gymnasium <2.0,>=0.29.1
- numpy <3.0,>=1.20
- torch <3.0,>=2.8
23 optional dependencies
- ale-py[extra]
- black[tests]
- matplotlib[extra]
- mypy[tests]
- myst-parser[docs]
- opencv-python[extra]
- pandas[extra]
- pillow[extra]
- psutil[extra]
- pygame-ce[extra]
- pytest[tests]
- pytest-cov[tests]
- pytest-env[tests]
- pytest-xdist[tests]
- rich[extra]
- ruff[tests]
- sphinx[docs]
- sphinx-autobuild[docs]
- sphinx-copybutton[docs]
- sphinx-rtd-theme[docs]
- sphinxcontrib-spelling[docs]
- tensorboard[extra]
- tqdm[extra]
Used By
- sb3-contrib
- datenwissenschaften
- csle-common
- infralib
- csle-agents
- sbx-rl
- godot-rl
- eta-incerto
- ants-ai
- pyforge3d
- eta-mistral
- deepmost
- cleanrl
- async-gym-agents
- reinforcement-learning-framework
- kitagentsdk
- elscirl
- il-datasets
- mouse-env
- sb3-extra-buffers
- gym-flp
- isopro
- ptan
- lineflow-rl
- ns-gym
- huggingbutt
- gr-libs
- scml-agents
- master-agent
- assetallocator
- attacks-on-drl
- commonpower
- swarm-drone-gym
- crosslearn
- slim-gym
- quantummetagpt
- loopquest
- neatrl
- dsmc-tool
- eagerx-dcsc-setups
- epyt-control
- pyhyflex
- ntn-torch
- gymnasium-search-race
- marlos
- cfrl
- rnl
- nett-benchmarks
- rlxbt
- pyrewardmachines