This technical report documents the development of novel Layout Analysis models integrated into the Docling document-conversion pipeline. We trained several state-of-the-art object detectors based on the RT-DETR, RT-DETRv2 and DFINE architectures on a heterogeneous corpus of 150,000 documents (both openly available and proprietary). Post-processing steps were applied to the raw detections to make them more applicable to the document conversion task. We evaluated the effectiveness of the layout analysis on various document benchmarks using different methodologies while also measuring the runtime performance across different environments (CPU, Nvidia and Apple GPUs). We introduce five new document layout models achieving 20.6% - 23.9% mAP improvement over Docling's previous baseline, with comparable or better runtime. Our best model, "heron-101", attains 78% mAP with 28 ms/image inference time on a single NVIDIA A100 GPU. Extensive quantitative and qualitative experiments establish best practices for training, evaluating, and deploying document-layout detectors, providing actionable guidance for the document conversion community. All trained checkpoints, code, and documentation are released under a permissive license on HuggingFace.
@article{arxiv.2509.11720,
title = {Advanced Layout Analysis Models for Docling},
author = {Nikolaos Livathinos and Christoph Auer and Ahmed Nassar and Rafael Teixeira de Lima and Maksym Lysak and Brown Ebouky and Cesar Berrospi and Michele Dolfi and Panagiotis Vagenas and Matteo Omenetti and Kasper Dinkla and Yusik Kim and Valery Weber and Lucas Morin and Ingmar Meijer and Viktor Kuropiatnyk and Tim Strohmeyer and A. Said Gurbuz and Peter W. J. Staar},
journal= {arXiv preprint arXiv:2509.11720},
year = {2025}
}
Comments
11 pages. 4 figures. Technical report for the layout models of Docling