Benchmark · reasoning

Humanity's Last Exam

Frontier of expert-level questions across domains.

1 hasil 1 model
0 24 48 72 96 2026-07-04 Theoria · 91.4 · 2026-07-04
Theoria
Linimasa
  1. 2026-07-04 Theoria 91.4% Theoria memverifikasi penalaran informal dengan mengaudit transisi keadaan bertipe