Research

Paper

AI LLM March 03, 2026

OCR or Not? Rethinking Document Information Extraction in the MLLMs Era with Real-World Large-Scale Datasets

Authors

Jiyuan Shen, Peiyue Yuan, Atin Ghosh, Yifan Mai, Daniel Dahlmeier

Abstract

Multimodal Large Language Models (MLLMs) enhance the potential of natural language processing. However, their actual impact on document information extraction remains unclear. In particular, it is unclear whether an MLLM-only pipeline--while simpler--can truly match the performance of traditional OCR+MLLM setups. In this paper, we conduct a large-scale benchmarking study that evaluates various out-of-the-box MLLMs on business-document information extraction. To examine and explore failure modes, we propose an automated hierarchical error analysis framework that leverages large language models (LLMs) to diagnose error patterns systematically. Our findings suggest that OCR may not be necessary for powerful MLLMs, as image-only input can achieve comparable performance to OCR-enhanced approaches. Moreover, we demonstrate that carefully designed schema, exemplars, and instructions can further enhance MLLMs performance. We hope this work can offer practical guidance and valuable insight for advancing document information extraction.

Metadata

arXiv ID: 2603.02789
Provider: ARXIV
Primary Category: cs.CL
Published: 2026-03-03
Fetched: 2026-03-04 03:41

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.02789v1</id>\n    <title>OCR or Not? Rethinking Document Information Extraction in the MLLMs Era with Real-World Large-Scale Datasets</title>\n    <updated>2026-03-03T09:26:40Z</updated>\n    <link href='https://arxiv.org/abs/2603.02789v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.02789v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Multimodal Large Language Models (MLLMs) enhance the potential of natural language processing. However, their actual impact on document information extraction remains unclear. In particular, it is unclear whether an MLLM-only pipeline--while simpler--can truly match the performance of traditional OCR+MLLM setups. In this paper, we conduct a large-scale benchmarking study that evaluates various out-of-the-box MLLMs on business-document information extraction. To examine and explore failure modes, we propose an automated hierarchical error analysis framework that leverages large language models (LLMs) to diagnose error patterns systematically. Our findings suggest that OCR may not be necessary for powerful MLLMs, as image-only input can achieve comparable performance to OCR-enhanced approaches. Moreover, we demonstrate that carefully designed schema, exemplars, and instructions can further enhance MLLMs performance. We hope this work can offer practical guidance and valuable insight for advancing document information extraction.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-03-03T09:26:40Z</published>\n    <arxiv:primary_category term='cs.CL'/>\n    <author>\n      <name>Jiyuan Shen</name>\n    </author>\n    <author>\n      <name>Peiyue Yuan</name>\n    </author>\n    <author>\n      <name>Atin Ghosh</name>\n    </author>\n    <author>\n      <name>Yifan Mai</name>\n    </author>\n    <author>\n      <name>Daniel Dahlmeier</name>\n    </author>\n  </entry>"
}