Project AURIONAI Use-Cases Research & Innovation

OmniDocBench

Document parsing overall score, higher is better

v1.6multimodalactive
9scores
9models
0independent
0maintainer-run
9vendor-reported

Definition

Version
v1.6
Category
multimodal
Maintainer
not in the notes
Status
active
Home page
not in the notes

Sources

  1. arxiv.org/pdf/2606.03264 https://arxiv.org/pdf/2606.03264

Scores at a Glance

Score, PaddleOCR-VL-1.6 paper table

Higher is better for this unit.

1BaiduPaddleOCR-VL-1.6Baiduoverall
96.33 score
vendor-reported2OpenDataLabMinerU2.5-ProOpenDataLaboverall
95.75 score
vendor-reported3Z.ai (Zhipu)GLM-OCRZ.ai (Zhipu)overall
95.22 score
vendor-reported4BaiduPaddleOCR-VL-1.5Baiduoverall
94.93 score
vendor-reported5OpenDataLabMinerU2.5 (MinerU-2.5)OpenDataLaboverall
93.04 score
vendor-reported6Google DeepMindGemini 3 ProGoogle DeepMind, Googleoverall
92.91 score
vendor-reported7GoogleGemini 3 FlashGoogleoverall
92.62 score
vendor-reported8rednote (Xiaohongshu)dots.ocrrednote (Xiaohongshu)overall
90.77 score
vendor-reported9DeepSeekDeepSeek-OCR 2DeepSeekoverall
90.25 score
vendor-reported
0255075100
Where this comes from
Each bar is one model, with its most independent recorded score. Hatched bars are vendor-reported; the word beside each bar says who ran it. Settings and dates differ between rows; the Scores table below has every result.

Scores

9 of 9 scores shown omnidocbench-v1-6-scores.csv
ModelScoreWho ran itSettingView or populationDateSource
BaiduPaddleOCR-VL-1.696.33 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
OpenDataLabMinerU2.5-Pro95.75 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
Z.ai (Zhipu)GLM-OCR95.22 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
BaiduPaddleOCR-VL-1.594.93 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
OpenDataLabMinerU2.5 (MinerU-2.5)93.04 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
Google DeepMindGemini 3 Pro92.91 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
GoogleGemini 3 Flash92.62 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
rednote (Xiaohongshu)dots.ocr90.77 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
DeepSeekDeepSeek-OCR 290.25 scorevendor-reportedoverallPaddleOCR-VL-1.6 paper table2 Jun 2026arxiv.org
Where this comes from
Scores from different settings, harnesses or dates are not directly comparable; the row says which. Sorted by score, highest first, as a reading aid, not a ranking.

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "OmniDocBench", https://aurion.semai.id/benchmarks/omnidocbench-v1-6/, records of October 2026.