khoj-ai/llm-coup

Let LLMs play coup with each other and see who's the best at deception & strategy

36
/ 100
Emerging

This project helps AI researchers and developers evaluate how well different large language models (LLMs) perform in situations requiring deception and complex strategy. By simulating games of Coup between various LLMs, it provides insights into their ability to bluff, strategize, and adapt. You input your chosen LLMs and desired game parameters, and it outputs game logs, results, and performance metrics for each model.

No commits in the last 6 months.

Use this if you need to systematically test and compare the strategic thinking and deceptive capabilities of different large language models in a controlled, game-theory-driven environment.

Not ideal if you're looking to play a game of Coup against an LLM, or if your primary interest is in evaluating LLMs on tasks unrelated to strategic interaction and deception.

AI-evaluation LLM-benchmarking game-theory-AI strategic-AI deception-modeling
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 4 / 25
Maturity 15 / 25
Community 15 / 25

How are scores calculated?

Stars

8

Forks

4

Language

TypeScript

License

GPL-3.0

Last pushed

Aug 18, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/khoj-ai/llm-coup"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.