BenchmarkActive
data quality57
Updated 15 h ago · first seen 11 Sept 2026
- Metric
- — · % ↑
- Current results
- 0
- Models
- 0
- Current leader
- o3 36.0%
No results recorded for this benchmark yet — its sources are being connected. The definition, aliases and variants are kept so links resolve; nothing is fabricated.
Leaderboard 4 models
Select models with +, then Compare.
4 results
One row per canonical model — its best current row inside this comparability group (effort variants are folded into the model). Bars are relative to the page's best score. “vs leader” reads comparability: partially comparable = same task, conditions differ (reasoning effort, temperature, judge). Rules →