Skip to content

Benchmark

Video-MME (long, no subtitles)

Vision benchmarks measure how well models understand images. Scores are percentages.

Models measured
1
Scores recorded
1
Sources
1
IQ weight
No weight published

Score distribution

72%
Mean
72%
Median
0%
Std dev
72%
10th percentile
72%
90th percentile
72%
Min

Leaderboard

#ModelScore
1GPT 4.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim72%

Data credits: Data by LLM Stats