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
Agents
23 models, higher is better
OpenAI's benchmark for browsing agents: questions whose answers are hard to find and need persistent searching across many web pages.
Benchmark data from llmmetric.com, Respan's model data service.