Intelligent Document Processing Leaderboard

A unified leaderboard for OCR, KIE, classification, QA, table extraction, and confidence score evaluation

📚 Datasets 📄 Paper 📝 Details 💻 Code 🏆 Leaderboard ✉️ Subscribe Get notified by email when new models or datasets are added.
This work is sponsored by Nanonets.

About the Leaderboard

The Intelligent Document Processing (IDP) Leaderboard provides a comprehensive evaluation framework for assessing the capabilities of various AI models in document understanding and processing tasks. This benchmark covers seven critical aspects of document intelligence:

  • Key Information Extraction (KIE): Evaluating the ability to extract structured information from documents
  • Visual Question Answering (VQA): Testing comprehension of document content through questions
  • Optical Character Recognition (OCR): Measuring text recognition accuracy across different document types
  • Document Classification: Assessing categorization capabilities
  • Long Document Processing: Evaluating performance on lengthy documents
  • Table Extraction: Testing tabular data understanding and extraction
  • Confidence Score: Measuring the reliability and calibration of model predictions

This benchmark is included in the Intelligent Document Processing (IDP) Leaderboard, which assesses the performance of different models in key information extraction tasks. For a comprehensive evaluation of document understanding tasks, please visit the full leaderboard.

Key Information Extraction (KIE) Leaderboard

Key Information Extraction (KIE) evaluates a model's ability to identify and extract specific information from documents, such as names, dates, amounts, and other structured data. This task measures how accurately models can locate and understand key entities within documents.

Rank Model Avg Nanonets-KIE Docile Handwritten-Forms
1 gemini-2.5-pro-preview-03-25 (reasoning: low) 79.66 91.00 65.79 82.18
2 qwen2.5-vl-32b-instruct 79.63 89.18 69.18 80.54
3 gemini-2.5-flash-preview-04-17 77.99 91.29 63.35 79.34
4 gemini-2.0-flash 77.22 88.31 65.06 78.28
5 qwen2.5-vl-72b-instruct 76.11 90.52 58.37 79.45
6 claude-3.7-sonnet (reasoning:low) 76.09 87.61 66.80 73.86
7 o4-mini-2025-04-16 75.43 86.91 59.52 79.85
8 mistral-medium-3 74.21 86.49 61.82 77.94
9 llama-4-maverick(400B-A17B) 73.30 85.78 61.70 72.43
10 gemma-3-27b-it 72.81 85.14 60.18 73.13
11 gpt-4.1-2025-04-14 72.68 87.85 61.20 68.98
12 claude-sonnet-4 71.91 85.78 63.53 66.42
13 gpt-4o-2024-08-06 71.83 88.63 56.37 70.48
14 gpt-4o-2024-11-20 70.91 88.03 56.56 68.15
15 InternVL3-38B-Instruct 70.31 84.02 57.47 69.42
16 gpt-4o-mini-2024-07-18 70.03 86.37 60.45 63.26
17 gpt-4.1-nano-2025-04-14 66.25 80.21 51.13 67.41
18 mistral-small-3.1-24b-instruct 63.73 75.47 47.07 68.64

BibTeX

@misc{IDPLeaderboard,
  title={IDPLeaderboard: A Unified Leaderboard for Intelligent Document Processing Tasks},
  author={Souvik Mandal and Nayancy Gupta and Ashish Talewar and Paras Ahuja and Prathamesh Juvatkar and Gourinath Banda},
  howpublished={https://idp-leaderboard.org},
  year={2025},
}
Souvik Mandal*1, Nayancy Gupta*2, Ashish Talewar*1, Paras Ahuja*1, Prathamesh Juvatkar*1,
Gourinath Banda*2
1Nanonets, 2IIT Indore