Benchmark
MVEB(text, video, v1)
Vision benchmarks measure how well models understand images. Scores are percentages.
- Models measured
- 4
- Scores recorded
- 4
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 39.2%
- Mean
- 48.4%
- Median
- 18.6%
- Std dev
- 10.4%
- 10th percentile
- 60.9%
- 90th percentile
- 10.4%
- Min