ocr-and-documents

You can extract text from scanned PDFs using OCR-based extraction via marker-pdf. It handles multi-language support and layout-aware outputs, turning unindexed files into machine-readable content for research data processing and indexing workflows.

The best way to convert a remote PDF URL to Markdown is using the web_extract path. It fetches online reports and research papers directly, extracting clean text and metadata for immediate ingestion into your knowledge base or downstream data pipeline.

Yes, you can use pymupdf for local text-based PDF extraction. It provides optional Markdown output and supports extracting tables and images, making it suitable for processing local research documents and archived reports without requiring OCR.

Yes, the tool handles complex layouts and figures using marker-pdf for advanced layout extraction. It processes research papers containing figures and references, outputting structured Markdown that preserves document layout for downstream data extraction.

You need OCR for document text extraction when working with scanned documents or image-based files that lack embedded text layers. The Skill automatically applies marker-pdf for OCR-enabled documents, while using pymupdf for standard text-based local PDFs.

Limitations of extracting text from archived PDFs depend on document quality and whether files are text-based or scanned. While pymupdf handles local text PDFs and marker-pdf applies OCR for scans, severely degraded scans may still produce imperfect layout extraction.