Open-Source LLMs: Punching Above Their Weight Class

Interactive analysis of 362 open-source models from Artificial Analysis. Evaluating peak reasoning performance across size tiers (Tiny, Small, Medium, Large).