Turn scanned pages into text and Markdown.
Extracts text and structured Markdown from images and PDFs using DeepSeek-OCR.
When to use it
Use when asked to extract image text, OCR a PDF, or convert a document to Markdown with DeepSeek-OCR.
Give it images or PDFs; it runs OCR and returns extracted text or Markdown.
What you provide
Requires CUDA 11.8 or newer.
Specifies Python 3.12.9, with conda recommended for setup.
This skill
GitHub
Downloads DeepSeek-OCR from GitHub
HuggingFace
Downloads the OCR model