Skip to content

Benchmark

OmniDocBench - Read Order Edit

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

Models measured
11
Scores recorded
11
Sources
1
IQ weight
0.03

Score distribution

15.9%
Mean
16.5%
Median
2.6%
Std dev
13%
10th percentile
19.3%
90th percentile
12.8%
Min
12.8%21.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.03 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Data credits: OmniDocBench (OpenDataLab), Apache-2.0 (licence)