Skip to content

Benchmark

MMEB Image - RefCOCO-Matching

Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.

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

Score distribution

91.1%
Mean
91.1%
Median
0%
Std dev
91.1%
10th percentile
91.1%
90th percentile
91.1%
Min

Leaderboard

#ModelScore
1Qwen3-VL Embedding 8BRefCOCO-Matching· MMEB (TIGER-Lab) data91.1%

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