MTG Bench: Testing How Well LLMs Can Play Magic: The Gathering

MTG Autodeck / Hacker News·medium signal

MTG Bench evaluates frontier LLMs on playing Magic: The Gathering, a game demanding long-horizon planning, hidden-information reasoning, and complex rule interactions. The benchmark drew 61 points and 31 comments on Hacker News. It probes reasoning weaknesses that standard knowledge benchmarks tend to miss.