Financial Statement Table Extractor
FinTabula is a modern, light, and beautiful web application built to automate the tedious process of extracting tabular data from financial statements (Balance Sheets, Income Statements, Cash Flow Statements) stored in PDF format. Built with FastAPI and a modern frontend, it bridges the gap between unstructured document data and structured analytical pipelines.
Dual Extraction Engines
FinTabula offers unparalleled flexibility through its dual-engine architecture:
- Local Engine: Powered by
pdfplumber, this engine parses vector-text PDF statements locally for instant, privacy-first, and zero-cost extraction. It precisely calculates bounding boxes and infers table structures based on strict coordinates. - Gemini AI Engine: Utilizing Google's Gemini 1.5 Flash multi-modal vision model, this engine intelligently scans and reconstructs complex, multi-page, or scanned (image-only) tables. It excels in semantic parsing, preventing word splits in dense columns, handling hierarchical headers, and perfectly mapping footnotes across margins.
Interactive Spreadsheet Editor
Beyond extraction, FinTabula features a sleek, dark-themed browser spreadsheet grid. It allows users to preview extracted sheets, double-click to edit cell values directly in the browser, add/remove rows and columns, and rename or delete sheets before generating the final export.
Premium Data Export
The application outputs beautifully styled Excel workbooks featuring zebra striping for readability, bold navy-filled headers, auto-fitted column widths, and intelligent cell alignment (numbers right, text left), ensuring the data is instantly presentation-ready.