/jump to repo

Repos Recent Bookmarks Watched Notes Tags Discover Compare Stats GitHub

Repository brief

openai/gym

Read the upstream summary on the left, browse the cached forks below it, and load each fork comparison into the right-hand panel.

Cached analysis

cached 2026-03-30T20:15:56.064Z

1mo ago

openai/gym

OpenAI Gym is a Python toolkit for developing and comparing reinforcement learning algorithms. It provides a standard API for environments, a set of compliant environments, and is widely used as a field standard, but the maintainers state future development has moved to Gymnasium and Gym will not receive future updates.

Loading tags...

Stars37,120

Forks8,706

Default branchmaster

Last pushed2026-03-26T23:13:27Z

Recommended shortcuts

Jump straight into Discofork's strongest cached fork picks, or open a compare view in one click.

Forks

Choose a fork to inspect

10 of 10 fork briefs

Maintenance:

Magnitude:

Sort:

Selected

Choose this fork only if Dart support is the main requirement. For general Gym adoption, upstream is newer and the fork is materially stale, so Gymnasium or modern Gym-compatible alternatives are the safer default.

Choose this fork if your goal is the more complex CarRacing variant and related experimentation. Choose upstream Gym or Gymnasium if you need a maintained, broadly compatible RL toolkit with the full general-purpose API surface.

Prefer this fork if your priority is RobotLocomotion/Drake simulation support and you want those environments embedded in a Gym-compatible package. Prefer upstream Gym or Gymnasium if you want broad compatibility, active maintenance, and minimal divergence.

Prefer this fork only if you already rely on its legacy customizations and are willing to own maintenance. For most adopters, upstream Gymnasium or current upstream-compatible alternatives are safer because this fork is both stale and materially diverged.

Prefer this fork only if you need its legacy demonstration/replay and scoreboard-oriented changes. For most adopters, upstream or Gymnasium is the safer choice because this fork is old, materially divergent, and missing later RL environment and vectorization work.

Adopt this fork only if you specifically need its older customized behavior or scoreboard/headless setup. For most users, upstream Gymnasium or a current Gym-compatible stack is the safer choice because this fork is stale and materially diverged.

Choose this fork only if you need a legacy Gym snapshot with local runtime tweaks and are prepared to own the maintenance burden. If you want current Gym behavior, broader compatibility, or easier long-term support, upstream Gymnasium is the better path.

Prefer this fork only if you need its legacy behavior or custom rendering/workflow changes and can accept substantial staleness. For new projects or active maintenance, upstream Gymnasium or a current fork is the safer default.

Prefer this fork only if you explicitly need legacy Gym behavior or are maintaining code tied to this old snapshot. For new work, upstream Gymnasium or a current Gym-compatible stack is the safer choice.

Choose this fork only if you need legacy Gym behavior or its custom headless/runtime changes. For new work, upstream Gymnasium or current Gym is the safer default because this fork is stale and materially behind upstream.

Fork comparison

DartEnv/dart-env

38/100

stale

significant_divergence

Choose this fork only if Dart support is the main requirement. For general Gym adoption, upstream is newer and the fork is materially stale, so Gymnasium or modern Gym-compatible alternatives are the safer default.

Likely purpose

To let adopters work with Dart-backed simulation environments inside the Gym API, likely for research or legacy projects that need those environments more than upstream parity.

Best for

Researchers or engineers who already depend on Dart-based RL environments.; Teams maintaining an old Gym-based codebase that needs the fork’s Dart additions.; Users prioritizing specific simulation content over upstream freshness.

Additional features

Adds a Dart environment area under `gym/envs/dart/` with large simulation assets and model files.
Expands example, docs, and environment test coverage around the Dart-specific stack.
Provides a forked RL toolkit path for users who need Dart support while keeping the Gym-style API.
Changes agent workflow behavior, such as prompt handling, handoffs, worktrees, memory, or orchestration flow.

Missing features

It lags upstream by 200 commits and has not been actively pushed since 2020-10-20, so newer upstream fixes and improvements are likely missing.
It likely misses later Gym maintenance and migration guidance reflected in upstream’s move to Gymnasium.
Upstream environment and tooling updates after the fork point, including newer compatibility and CI changes, are not present here.
Some upstream behavior appears to be removed, disabled, or intentionally stripped down in this fork.
It trails upstream by 200 commits, so some recent upstream features and fixes are likely not present yet.

Strengths

Useful if you specifically need Dart environments inside the Gym ecosystem.
Large amount of Dart asset work suggests the fork is purpose-built rather than a shallow rebrand.
Can be a practical legacy base for existing code that already targets Gym semantics.

Risks

Stale maintenance makes it a poor choice for new projects that need current Gym behavior or security/compatibility fixes.
Substantial divergence raises integration risk when trying to merge with modern upstream Gym or Gymnasium.
The fork’s value is narrow if you do not need Dart support.