Artificial Analysis: New AI Benchmarks for Professional Domains

Artificial Analysis has launched six new capability indices to benchmark AI models across professional domains. These indices evaluate frontier AI models in fields like finance, legal, healthcare, strategy, engineering, and economics. What's interesting is how these indices measure performance. They are composite benchmarks, blending multiple evaluation metrics into a single score for each domain. For example, the Legal Index weights Legal Knowledge at 35% and Agentic Knowledge Work at 25%. Early results show variants of Claude Fable 5 and Claude Opus 4.8 performing well on the Legal Index. This signals a shift toward more domain-specific AI assessment. The bottom line is these benchmarks are designed to simulate real-world tasks professionals perform, offering a more practical evaluation of AI capabilities.