Extracts structured text from images and scanned documents using multiple high-performance OCR engines.
The LlamaFarm OCR skill empowers Claude to convert visual data from images and scanned PDFs into actionable, machine-readable text directly within your development environment. By supporting specialized backends like Surya, PaddleOCR, and EasyOCR, it allows users to accurately process everything from standard business documents and CJK characters to handwritten notes and scene text. This skill streamlines document processing workflows by providing layout information and confidence scores, making it an essential tool for digitizing legacy data, indexing documents for RAG, or automating data entry tasks.
主な機能
01Seamless integration with downstream tasks like classification and entity extraction
02Processes common image formats and multi-page scanned PDF documents
03Extensive multi-language support covering 100+ languages including CJK and handwriting
04Supports multiple backends including Surya, EasyOCR, PaddleOCR, and Tesseract
05Detailed output including extraction confidence scores and layout metadata
060 GitHub stars
ユースケース
01Converting screenshots of terminal output or UI designs into editable text for documentation
02Extracting content from scanned research papers to build local RAG knowledge bases
03Digitizing paper invoices and receipts for automated data entry and accounting
