Skip to content

Benchmark

MMEB VisDoc - VisRAG_ChartQA

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

88.3%
Mean
88.3%
Median
0%
Std dev
88.3%
10th percentile
88.3%
90th percentile
88.3%
Min

Leaderboard

#ModelScore
1Qwen3-VL Embedding 8BVisRAG_ChartQA· MMEB (TIGER-Lab) data88.3%

Data credits: MMEB Leaderboard (TIGER-Lab), Apache-2.0 (licence)