Dicklesworthstone/llm_aided_ocr
Enhances Tesseract OCR output using LLMs (local or API) for error correction, smart chunking, and markdown formatting of scanned PDFs
This project helps anyone working with scanned documents by taking a PDF and transforming its raw text into a much more accurate, readable, and well-formatted document. It cleans up common scanning errors, organizes the content into logical sections, and can even format it using Markdown. This is ideal for researchers, analysts, or anyone who frequently extracts information from old reports, books, or archived files.
2,885 stars. Actively maintained with 2 commits in the last 30 days.
Use this if you need to reliably extract clean, structured text from scanned PDFs, especially those with quality issues or complex layouts.
Not ideal if you primarily work with born-digital PDFs or already have perfectly clean text, as the overhead of advanced correction may not be necessary.
Stars
2,885
Forks
206
Language
Python
License
—
Category
Last pushed
Mar 03, 2026
Commits (30d)
2
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/Dicklesworthstone/llm_aided_ocr"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
NanoNets/docstrange
Extract and convert data from any document, images, pdfs, word doc, ppt or URL into multiple...
th1nhhdk/local_ai_ocr
An local, offline (after initial setup), portable OCR software that can process images and PDF...
emcf/thepipe
Get clean data from tricky documents, powered by vision-language models ⚡
langstruct-ai/langstruct
Extract structured data from any content using LLMs.
hashangit/Extract2MD
Extract2MD is a powerful and versatile AI-enabled client-side JavaScript library for extracting...