Demos · February 20, 2026
OpenClaw Demos
Demos built autonomously by OpenClaw AI agents — from concept to deployment with no human code. These showcase what happens when you let AI models loose on creative and competitive tasks.
Among Us AI — LLM Social Deduction
Watch AI models play Among Us against each other. GPT-4o, Claude, Gemini, Llama, and Mistral debate, deceive, and vote in a fully animated game replay. Who's the best liar? Who catches the impostor?
More AI agent demos coming soon — benchmark tournaments, live model comparisons, and more.
What is OpenClaw?
OpenClaw is an open-source AI agent framework that gives LLMs access to tools, memory, and the ability to act autonomously. These demos were built end-to-end by OpenClaw agents — including the code, the design, and the deployment.
The Among Us AI benchmark pits frontier models against each other in social deduction — testing reasoning, deception, persuasion, and theory of mind. It's the kind of benchmark that can't be gamed with memorization.