Original
Duplicate
Track the Epstein files* released by the US DOJ on January 30, 2026
*PDF documents only - images & video excluded.
Entity extraction, redaction detection, and removal tracking — tools the official DOJ site doesn't offer.
Removal tracking
Files have been taken down. We saved them on day one.
We archived files within hours of the initial DOJ release. Since then, files have been quietly removed. Our heatmaps track exactly which documents disappeared and when.
Browse datasets →
Redaction detection
Documents have been altered. See exactly what changed.
Pixel-level comparison across duplicate files reveals redacted text, swapped pages, and subtle alterations invisible to casual readers. Side-by-side diff viewing shows every change.
Explore redactions →
Entity search
Search with precision the official site cannot match.
Every page is run through named-entity recognition and regex extraction. Search by person, phone number, email, date, or account number — with exact page references and PDF preview.
Try entity search →
AI classification
1.4 million PDF files with no context? We're fixing that.
Every page in PDF is classified as text-based or image-based. Text pages are run through OCR, normalized entities extracted, and brief summary added by AI. Image pages are captioned by AI. Documents flagged as likely to contain CSAM are excluded from preview.