# NVIDIA-NeMo/Gym

Evaluate and improve models and agents using environments

Repository: https://github.com/NVIDIA-NeMo/Gym
Canonical: https://ross.abutalabs.com/products/nvidia-nemo-gym
Homepage: https://docs.nvidia.com/nemo/gym/main/about/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: reinforcement-learning, reinforcement-learning-environments, rl-environment, rl-training, gym, agents, benchmarks, environments, evaluation, llm
Last push: 2026-09-03T01:59:55+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 84, longevity 26
- inputs: {"age_days": 373, "days_push": 0, "days_rel": 27, "gap_med": 35.0, "n_releases_24m": 7}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1156, forks 304 (observed 2026-09-03T02:15:06.762483+00:00)

## What it is
NeMo Gym is a Python library from NVIDIA for evaluating and improving LLM models and agents using environments. It provides infrastructure to build environments (dataset, agent harness, verifier, state), run evaluation and RL training at scale, and ships a hub of popular benchmarks and training environments.

## Use cases
- evaluate llm agents in stateful environments
- run rl training environments for language models
- benchmark models on tool calling and code execution tasks
- build custom environments with verifiers for agent training
- scale agent evaluation to thousands of concurrent rollouts
- integrate rl frameworks with agent environments
- reproducible agent benchmarking across teams

## When to choose
- you need stateful evaluation of models or agents with verifiers
- you want to scale evaluation or training to thousands of concurrent environments
- you want to transition seamlessly between evaluation, agent optimization, and RL training
- you want a shared environment hub of popular benchmarks

## When to avoid
- you just need a stateless check on model outputs - a simple script suffices
- you need a general-purpose game RL gym rather than LLM agent environments
- you are not working with LLMs or agents

## Facets
- artifact type: library
- maturity: active
- function: reinforcement-learning, agent-framework, benchmarking, llm-training, testing, mcp
- domain: reinforcement-learning, large-language-models, machine-learning, developer-tools
- platform: python, cli
- tags: rl-environments, agent-evaluation, verifiers, llm-benchmarks, nvidia-nemo, agent-harnesses, sandboxing, ai-agents, linux, docker, gpu

## Member repositories
- NVIDIA-NeMo/Gym (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:06.762483+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:35:22.485708+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA-NeMo/Gym (fetched 2026-09-03T02:15:06.762483+00:00, sha e012551caf47)
  - homepage: https://docs.nvidia.com/nemo/gym/main/about/ (fetched 2026-08-29T12:40:30.425217+00:00, sha d311c13db343)
  - site_page: https://docs.nvidia.com/nemo/gym/about (fetched 2026-08-29T12:40:30.434377+00:00, sha 02a7e0adafc8)
  - site_page: https://docs.nvidia.com/nemo/gym/main/about/concepts (fetched 2026-08-29T12:40:30.436035+00:00, sha dd71c5b93d30)
  - site_page: https://docs.nvidia.com/nemo/gym/main/about/architecture (fetched 2026-08-29T12:40:30.437855+00:00, sha bb9efe563e93)
  - site_page: https://docs.nvidia.com/nemo/gym/main/about/ecosystem (fetched 2026-08-29T12:40:30.439698+00:00, sha bae5d4c68fdd)
  - site_page: https://docs.nvidia.com/nemo/gym/main/about/release-notes (fetched 2026-08-29T12:40:30.441381+00:00, sha 5dbbc4111c59)
  - site_page: https://docs.nvidia.com/nemo/gym/main/get-started/installation (fetched 2026-08-29T12:40:30.443449+00:00, sha ad679ca7a8d7)
  - site_page: https://docs.nvidia.com/nemo/gym/main/get-started/quickstart (fetched 2026-08-29T12:40:30.445018+00:00, sha 4d4a6c29a2f6)
  - site_page: https://docs.nvidia.com/nemo/gym/main/reference/faq (fetched 2026-08-29T12:40:30.446663+00:00, sha 8e5a41ae4c75)
- Data as of 2026-08-30T08:39:29.467469+00:00.