Every model's score on every benchmark we track, each linked to where it was published. Independent runs and lab-reported numbers are kept apart. Many of these benchmarks feed the Respan Index.
177 of 177 benchmarks
Reasoning
17 models, higher is better
Graduate-level, 'Google-proof' multiple-choice questions in biology, physics, and chemistry written by domain experts to be hard to answer even with web search.
Benchmark data from llmmetric.com, Respan's model data service.