Independent AI benchmark evidence

Find the benchmark behind every score.

Search benchmark definitions, categories, measurement types, reporting sources, and model results. Every result stays linked to dated source evidence.

Current snapshot

Evidence coverage

Loading dashboard snapshot…

Loading coverage overview…

Browse by capability

Benchmark categories

Choose a category to open the benchmark directory with that filter applied.

Loading categories…

Manufacturer catalog

Latest manufacturer releases

Recent manufacturer-verified releases, ordered independently of benchmark scores.

Loading latest releases…

Current leaderboard

Comparable capability evidence

Eligible models have a verified release on or after 2025-09-01 and evidence from at least three normalized benchmarks. Release recency does not affect ranking.

Loading comparable evidence…

? Data & methodology

Loading methodology…