A benchmark is a standardized test used to compare AI systems: fixed tasks with fixed scoring, so results are comparable across models. Popular examples test reasoning, coding, math, or knowledge — but every benchmark measures only what its tasks actually capture.
Benchmark scores drive headlines and marketing. Knowing their limits — saturation, data contamination, narrow scope — helps you read 'state of the art' claims with healthy skepticism.
When two labs both claim the 'best' model, the benchmark tables usually reveal they each chose the test where their model wins.