ModelRadar
/
← All benchmarks

Video-MME

multimodal · video-understanding↗ Quelle
81Evidence
active

What it measures

Videoverstaendnis ueber kurze bis sehr lange Clips, mit und ohne Untertitel.

What it does not measure

Standbilder, Audio allein.

How the score is composed

Currency82
Discrimination74
Headroom100
Contamination safety55
Gaming resistance100
Methodological rigor80
Reproducibility60

Details

Benchmaxxing risk
low
Difficulty
high
Public test set
yes
Human-verified
yes
LLM-judged
no
Letzter Messwert
2026-07-25
Skala
0–100 % correct (höher besser)

Best measured models

  1. 1Mage-VLMicrosoftvendor number59.7 % correct
  2. 2Phi-4-multimodalMicrosoftvendor number50.8 % correct