# docx-smart-extractor

Extract and analyze Word documents (1MB-50MB+) with minimal token usage. Lossless extraction of all text, tables, formatting, and document structure while achieving 10-50x token reduction through local extraction, semantic chunking by headings, and intelligent caching.