Benchmark leaderboard
Humanity's Last Exam is a closed-ended set of expert-written questions across dozens of academic fields, built to stay hard after older exams saturated. Artificial Analysis runs it independently with the same prompting and grading for every model. This page lists every model ModelCap tracks with a published AA Humanity's Last Exam result, its best evaluated configuration, the source's own rank, and the model's ModelCap Index position, API price and context window.
Snapshot as of 11 August 2026
Sorted by published score, one row per model. Source: Artificial Analysis.
No tracked model currently publishes a AA Humanity's Last Exam result in this snapshot. The standings return with the next dataset refresh that carries one; the source board itself stays available at the link above.
AA Humanity's Last Exam ranks models by Humanity's Last Exam accuracy. ModelCap ingests the board as published, matches each entry to a catalogue model with a reviewed identity, and shows the source's score, interval and rank unchanged. Where a model has an admitted result, it feeds the ModelCap Index as reasoning evidence alongside the other public boards; the methodology documents the weighting and the identity rules.
Humanity's Last Exam is a closed-ended set of expert-written questions across dozens of academic fields, built to stay hard after older exams saturated. Artificial Analysis runs it independently with the same prompting and grading for every model. It is published by Artificial Analysis.
No tracked model currently publishes a AA Humanity's Last Exam score.
None in this snapshot: no tracked model has an admitted AA Humanity's Last Exam result right now. The board repopulates from the sealed dataset as soon as one is admitted.
The board ranks models by Humanity's Last Exam accuracy; higher scores are better. ModelCap shows the source's own score, interval and rank and never re-runs the evaluation.
Not on its own. The ModelCap Index combines several public capability sources with published uncertainty; AA Humanity's Last Exam contributes as reasoning evidence where a model has an admitted result. The methodology page documents the weighting.
ModelCap holds no AA Humanity's Last Exam publication in this snapshot. The page re-renders every minute from the sealed dataset and repopulates with the next admitted source row.