GeneBench-Pro — Mercor evaluation (GeneBench-Pro (Mercor))

We captured Mercor's GeneBench-Pro results on November 12, 2026. The table preserves 36 published configurations, their source identifiers, reasoning settings, and disclosed mean score values. These are results from Mercor's evaluation setup; we did not rerun them.

This dated source table is display only and does not enter overall or category rankings. Source configurations retain their external identity, including Pro descriptors and different effort settings.

Mercor scores the 10-problem public set of OpenAI's GeneBench-Pro, with ten runs per model on each problem, so margins are wide. The scores labs report on their own runs of the benchmark are listed separately and are not combined with this table.

The row labels retain published error margins and sample counts when supplied. Missing margins and counts remain unreported; a sample count is not used to reconstruct a task denominator. Except where the dedicated source defines an interval, the confidence level is unreported. Equal displayed scores and overlapping margins do not establish statistical ties. Capture time does not establish individual evaluation dates.

The general methodology lists binary: the final answer is checked against a recoverable target within calibrated tolerances and a judge of Programatic answer checking. Its run-count label is k = 10. The linked result page takes precedence where its setup differs, as noted above.