Tensorforce provides a TensorFlow library for applied reinforcement learning tasks.
Open source
24 alternativesGymnasium
Gymnasium provides a standard API for single agent reinforcement learning environments and related utilities.
About Gymnasium
Gymnasium is an open source Python library for developing and comparing reinforcement learning algorithms. It provides a standard API for communication between learning algorithms and environments, along with a standard set of environments that follow that API. It is a fork of OpenAI's Gym library, maintained by Gym's maintainers, and future maintenance will take place in this repository. Its scope is single-agent reinforcement learning.
The library includes environment families such as Classic Control, Box2D, Toy Text, MuJoCo, and Atari, and it is designed to work with third-party environments that follow the same API. The base installation does not include dependencies for every environment family. Some of these dependencies can be difficult to install on certain systems, so specific families must be installed separately.
Key features
- Standard API connecting learning algorithms with reinforcement learning environments
- Standard set of environments that follow the same API
- Environment families including Classic Control, Box2D, Toy Text, MuJoCo, and Atari
- Optional installation of dependencies for individual environment families
- Version suffixes on environment IDs that change when updates may affect learning results
- apply_env_compatibility option in gymnasium.make for some third-party environments
Best for: Researchers and developers who build or compare single-agent reinforcement learning algorithms in Python.
Good for
- Provides a common interface so algorithms and environments can be developed and compared consistently
- Maintained by the team that took over Gym, with future maintenance occurring in this repository
- Toy Text environments are small and suitable for debugging reinforcement learning implementations
Watch out for
- The base installation does not include dependencies for all environment families
- Some environment dependencies can be problematic to install on certain systems
- Some third-party environments were built for specific versions and may need compatibility handling
Gymnasium at a glance
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
Gymnasium alternatives
Tools and sites you could use instead, ranked by how closely they match.
DyNet provides tools for building dynamic neural networks and deep learning models.
Open source
24 alternatives
MII makes low-latency and high-throughput inference possible, powered by DeepSpeed.
Open source
24 alternatives
The colibri-core NLP tool and library works with n-grams, skipgrams, and other linguistic patterns.
Open source
24 alternatives
DEAP provides a Python framework for prototyping evolutionary computation algorithms and testing new ideas.
Open source
24 alternatives
Extensible Toolkit for Finetuning and Inference of Large Foundation Models.
Open source
24 alternativesPopular in AI Infrastructure & MLOps
Well-known listings from the same topic.
Guild AIguild.aiFlow AIflow-ai.comsystemprompt.iosystemprompt.ioCoreWeavecoreweave.comKili Technologykili-technology.comAPI docsdocs.x.ai
Gymnasium FAQ
How do I install the base Gymnasium library?
Use pip install gymnasium. This does not include dependencies for every environment family.
How do I install dependencies for all environment families?
Use pip install "gymnasium[all]". To install one family, use a command such as pip install "gymnasium[atari]".
What do the version suffixes on environment names mean?
Environment names end in a suffix such as -v0. The number increases when changes might affect learning results, to prevent confusion and support reproducibility.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.