Request access to SecureFinAI Contest 2025 · Task 3 English OCR

SecureFinAI Contest 2025 · Task 3 English OCR is released by The Fin AI for research. Access is granted automatically after you complete this short form.

By accessing this dataset you agree to its license and to cite the The Fin AI paper in any resulting publication.

Log in or Sign Up to review the conditions and access this dataset content.

SecureFinAI Contest 2025 · Task 3 English OCR

🌐 The Fin AI

Formerly TheFinAI/SecureFinAI_Contest_2025-Task_3_EnglishOCR (the old name redirects here).

Images of English financial documents with their text and matched HTML, released for Task 3 of the SecureFinAI Contest 2025. Some documents are public filings that contain names of individuals. The authors have not yet specified a license for this dataset; treat it as research-only until clarified, and respect the terms of the original sources.

Task financial document OCR
Language en
License other (unspecified)

Quick Start

from datasets import load_dataset

ds = load_dataset("TheFinAI/en-securefinai-2025-ocr", split="test")
print(ds[0])
Split Rows
train 3,415
test 1,500

Columns: image, text, matched_html

Downloads last month
48

Models trained or fine-tuned on TheFinAI/en-securefinai-2025-ocr