junhoyeo/BetterOCR
🔍 Better text detection by combining multiple OCR engines (EasyOCR, Tesseract, and Pororo) with 🧠 LLM.
Struggling with inaccurate text extraction from images or documents, especially in non-English languages? This tool helps you get much more reliable text by combining several OCR (Optical Character Recognition) tools and then using AI to clean up and correct any errors. It takes an image or document as input and provides accurate, readable text, ideal for anyone who needs to digitize information from physical or scanned sources.
622 stars. No commits in the last 6 months.
Use this if you need to extract text from images, scans, or photos and are frequently frustrated by errors, especially with non-English content or noisy documents.
Not ideal if you require real-time, high-volume text extraction where latency is critical, as combining multiple engines and an LLM can take more time.
Stars
622
Forks
38
Language
Python
License
MIT
Category
Last pushed
Jun 10, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/junhoyeo/BetterOCR"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
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...
Dicklesworthstone/llm_aided_ocr
Enhances Tesseract OCR output using LLMs (local or API) for error correction, smart chunking,...
emcf/thepipe
Get clean data from tricky documents, powered by vision-language models ⚡
langstruct-ai/langstruct
Extract structured data from any content using LLMs.