Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
LegalBenchlegalbench.aiLegalBench is a collaborative benchmarking project that evaluates foundation models on legal reasoning tasks. It aims to determine what types of legal reasoning these models can perform and to develop effective prompting strategies, through collaboration between computer science and legal scholars. The project seeks to understand models' limitations and capabilities to enable safe, ethical applications of AI in the legal domain.
Parse Score
