benchgap
Calibration

VideoMMMU → Video-MME (with subtitle)

Video-MME (with subtitle) is estimated from VideoMMMU with a linear curve fitted on 5 models measured on both: y = 0.8178·x + 0.1781, R² = 0.74, cross-validated error 1.2 pp. It is used for 2 estimates.

Estimated modelVideoMMMUVideo-MME (with subtitle)Source
dots3-note Preview86.8%88.8%estimated ± 1.2 pp, medium confidence
Kimi K2.586.6%88.6%estimated ± 1.2 pp, medium confidence