Document Text Reader & Keyword Search

What the document reader actually does

This tool extracts text in your browser, displays selected sentences, and ranks passages by matching keywords from your query. It does not call a language model to reason about a document. A matching passage can help you locate information, but it is not a generated answer or a verified conclusion.

1. Choose a small sample document with selectable text.
2. Review extracted highlights against the original document.
3. Search for a distinctive name or phrase present in the text.
4. Open the cited page or section in the original to confirm context, units, and exceptions.

Know the extraction limits

Scanned image PDFs need OCR, which this tool does not provide. Complex tables, reading order, and legacy DOC extraction can produce incomplete text. Keyword matching can miss synonyms and does not calculate totals from a CSV. An absent match does not prove that information is absent from the source.

The reader code processes selected files in the browser without an explicit file upload request. The page also loads third-party scripts. Use a non-confidential sample and close or reload the page to clear the current document from this tool.

Guidance revised 22 October 2026.