← all datasets

OmniDocBench v-1.5

Emerging
1papers using it
2026first seen

OmniDocBench v-1.5 is a benchmark dataset designed to evaluate the performance of OCR models on complex document structures by providing high-quality structured data that addresses diverse layouts and rare document types.

🔎 Find this dataset

Papers using OmniDocBench v-1.5 (1)

OmniDocBench v-1.5 dataset — papers, benchmarks & downloads · Large Language Models