# Concordia

A Google DeepMind library for building generative social simulations, where language-model entities describe actions in natural language that a Game Master resolves into outcomes.

Concordia is a library for constructing and running generative agent-based models that simulate interactions among entities in grounded physical, social or digital environments, using an interaction pattern borrowed from tabletop role-playing games. A special entity called the Game Master simulates the environment: player entities describe their intended actions in natural language, and the Game Master translates those into outcomes, such as checking whether an action is physically plausible.

The library is organized around three concepts: entities, either player agents or Game Masters; components, modular building blocks such as memory systems, reasoning chains and sensory modules that define an entity's behavior; and an engine that runs the simulation loop, soliciting actions and delegating their resolution. Concordia supports applications including social-science research, AI safety and ethics, cognitive neuroscience, economics, synthetic data generation, and evaluating real services through simulated usage, and it requires access to a standard LLM API. It is written in Python and licensed under Apache 2.0.

Website: https://github.com/google-deepmind/concordia