Skip to content

Benchmark

MMEB Image - TextVQA

OCR benchmarks measure how well models read text in images and documents. Scores are percentages.

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

Score distribution

92.8%
Mean
92.8%
Median
0%
Std dev
92.8%
10th percentile
92.8%
90th percentile
92.8%
Min

Leaderboard

#ModelScore
1Qwen3-VL Embedding 8BTextVQA· MMEB (TIGER-Lab) data92.8%

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