← All benchmarks

GDPval Extended (Mercor held-out rubric score)

Professional Work

Mercor's held-out Extended twin of OpenAI's GDPval: 132 Mercor knowledge-work tasks graded by Mercor's rubric mean score. Like its public twin, Mercor's 213-task GDPval rubric score (gdpval-mercor-213-rubric, under gdpval), the rubric is undocumented and differs from GDPval's blind expert pairwise grading, so this is a diagnostic. Mercor does not publish the held-out tasks or say how they were drawn or written, so this is its own benchmark, not a component of the public one.

Top models (higher is better)