MODEL BENCHMARKS

How every AI model scores

Published results for language models across coding, math, reasoning, knowledge and agent benchmarks, plus separate boards for vision, image, video and audio models.