# openai/multiagent-particle-envs

Code for a multi-agent particle environment used in the paper "Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments"

Repository: https://github.com/openai/multiagent-particle-envs
Canonical: https://ross.abutalabs.com/products/multiagent-particle-envs
Homepage: https://arxiv.org/pdf/1706.02275.pdf
Language: Python
License: MIT
License Family: permissive
Topics: paper
Archived: true
Last push: 2024-04-09T01:15:17+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3303, "days_push": 877, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2774, forks 820 (observed 2026-08-28T04:07:19.145352+00:00)

## What it is
A simple multi-agent particle world environment with continuous observations and discrete actions, used in the paper 'Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments'. It provides Gym-like environments with customizable scenarios for multi-agent reinforcement learning research.

## Use cases
- train multi-agent reinforcement learning policies
- reproduce the MADDPG paper experiments
- create custom cooperative-competitive multi-agent scenarios
- visually inspect agent behaviors in a particle world
- benchmark multi-agent algorithms on simple environments

## When to choose
- reproducing the Multi-Agent Actor-Critic paper
- prototyping multi-agent RL algorithms on lightweight environments
- teaching multi-agent reinforcement learning concepts

## When to avoid
- you need maintained, pip-installable environments with modern Python support (use PettingZoo instead)
- you need photorealistic or complex 3D simulations
- you need production-ready software

## Facets
- artifact type: library
- maturity: abandoned
- function: simulation, machine-learning, reinforcement-learning
- domain: reinforcement-learning, machine-learning, simulation
- platform: python, cross-platform
- tags: multi-agent, gym-environments, particle-environment, research-code, archived, cooperative-competitive, marl, research

## Member repositories
- openai/multiagent-particle-envs (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:19.145352+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 2026-08-30T08:16:35.288678+00:00, confidence not recorded.
  - readme: https://github.com/openai/multiagent-particle-envs (fetched 2026-08-28T04:07:19.145352+00:00, sha 0557648375b4)
- Data as of 2026-08-30T08:39:29.467469+00:00.