Multi-Modal Vision vs. Text-Based Parsing: Benchmarking LLM Strategies for Invoice Processing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Berghaus, David, Berger, Armin, Hillebrand, Lars, Cvejoski, Kostadin, Sifa, Rafet
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911138730475520
author Berghaus, David
Berger, Armin
Hillebrand, Lars
Cvejoski, Kostadin
Sifa, Rafet
author_facet Berghaus, David
Berger, Armin
Hillebrand, Lars
Cvejoski, Kostadin
Sifa, Rafet
contents This paper benchmarks eight multi-modal large language models from three families (GPT-5, Gemini 2.5, and open-source Gemma 3) on three diverse openly available invoice document datasets using zero-shot prompting. We compare two processing strategies: direct image processing using multi-modal capabilities and a structured parsing approach converting documents to markdown first. Results show native image processing generally outperforms structured approaches, with performance varying across model types and document characteristics. This benchmark provides insights for selecting appropriate models and processing strategies for automated document systems. Our code is available online.
format Preprint
id arxiv_https___arxiv_org_abs_2509_04469
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multi-Modal Vision vs. Text-Based Parsing: Benchmarking LLM Strategies for Invoice Processing
Berghaus, David
Berger, Armin
Hillebrand, Lars
Cvejoski, Kostadin
Sifa, Rafet
Computation and Language
Artificial Intelligence
This paper benchmarks eight multi-modal large language models from three families (GPT-5, Gemini 2.5, and open-source Gemma 3) on three diverse openly available invoice document datasets using zero-shot prompting. We compare two processing strategies: direct image processing using multi-modal capabilities and a structured parsing approach converting documents to markdown first. Results show native image processing generally outperforms structured approaches, with performance varying across model types and document characteristics. This benchmark provides insights for selecting appropriate models and processing strategies for automated document systems. Our code is available online.
title Multi-Modal Vision vs. Text-Based Parsing: Benchmarking LLM Strategies for Invoice Processing
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2509.04469