Translated data: The New York-based AI startup Arthur has released an open-source tool called ArthurBench, designed for evaluating and comparing the performance of large language models. ArthurBench enables businesses to test the performance of various language models on specific use cases and provides metrics for comparison such as accuracy, readability, and risk avoidance. Financial services companies, vehicle manufacturers, and media platforms have already begun using ArthurBench, accelerating analysis and delivering more accurate answers.