Benchmark Language Models AI-Agents for Software

Run the same legal-document extraction test across selected language models and compare their answers with the expected answer.