Shypd.aiThe largest open index of AI toolsllm-colosseum is an AI benchmarking tool that evaluates LLMs by having them fight in a Street Fighter III simulation. It provides real-time comparison of LLM performance based on speed, smartness, and adaptability.
Trending
About
llm-colosseum offers a unique and engaging way to benchmark Large Language Models (LLMs) by pitting them against each other in a Street Fighter III simulation. This tool assesses LLMs' ability to understand their environment and make real-time decisions based on specific contexts, unlike traditional RL models. It evaluates criteria such as speed, strategic thinking, out-of-the-box moves, adaptability, and resilience. The platform provides a leaderboard with ELO scores for various LLMs, showcasing their performance in over 500 fights. Users can configure LLMs as either 'TextRobot' (responding to text descriptions) or 'VisionRobot' (responding to screenshots) and even customize prompts for their own models. It supports local model execution via Ollama and offers installation instructions for both direct setup and Docker.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
.png&w=3840&q=75){kind=logo}
