Can I use this to search for specific names or terms across a FOIA document dump?

Yes. Extract all PDFs using the combined output mode, then run any text search tool across the resulting file. This is significantly faster than opening each PDF individually in a viewer.

What if some documents in my FOIA response are scanned image PDFs?

Scanned PDFs produce empty text output. Government agencies frequently scan physical records before releasing them via FOIA, so a mixed batch of digital and scanned PDFs is common. Scanned documents need OCR processing before text extraction will work.

Will redacted text appear in the extracted output?

Properly redacted PDFs replace text with black boxes at the rendering layer — the underlying text is removed. If a redaction was applied only as a visual overlay, the text may still be present in the extraction output, but this is uncommon in properly processed government documents.

Is there a file size limit per document?

Each PDF can be up to 300 MB. Court filings and FOIA responses are typically under 50 MB per document, so most journalism research batches process without hitting size limits.

Extract Searchable Text from PDF Document Dumps for Investigative Research

Investigative journalists and researchers who receive FOIA responses, court filings, or leaked document sets as PDF bundles face a specific problem: hundreds or thousands of pages of content locked in a format that is slow to search and impossible to import into analysis tools without extraction. Deliteful converts up to 50 PDFs to searchable plain text simultaneously, turning a document dump into a searchable corpus in minutes.

Create free account

Public records requests frequently produce PDF responses ranging from a dozen to several hundred documents. Reviewing these manually is the only option when the text is locked in PDF format — but once extracted, the same corpus becomes searchable with any text tool. Journalists have used text extraction workflows to identify key names, dates, and phrases across large document sets without reading every page, a technique central to data-driven investigative reporting.

The combined output mode is especially powerful for document dump analysis: a single text file containing all extracted content allows you to run grep, use a desktop search tool, or paste into a document analysis environment. For court filings where each document needs to be cited individually, per-file output preserves document identity. Both modes process up to 50 PDFs per batch, with each file up to 300 MB.

How it works

1
Upload the document set
Add up to 50 PDFs from your FOIA response, court filing set, or document production.
2
Choose corpus or per-file output
Combined for full-set searching and analysis; per-file for source-traceable citation workflows.
3
Search and analyze
Download .txt files and run keyword searches, entity extraction, or timeline analysis across the full document set.

Frequently asked questions

Can I use this to search for specific names or terms across a FOIA document dump?: Yes. Extract all PDFs using the combined output mode, then run any text search tool across the resulting file. This is significantly faster than opening each PDF individually in a viewer.
What if some documents in my FOIA response are scanned image PDFs?: Scanned PDFs produce empty text output. Government agencies frequently scan physical records before releasing them via FOIA, so a mixed batch of digital and scanned PDFs is common. Scanned documents need OCR processing before text extraction will work.
Will redacted text appear in the extracted output?: Properly redacted PDFs replace text with black boxes at the rendering layer — the underlying text is removed. If a redaction was applied only as a visual overlay, the text may still be present in the extraction output, but this is uncommon in properly processed government documents.
Is there a file size limit per document?: Each PDF can be up to 300 MB. Court filings and FOIA responses are typically under 50 MB per document, so most journalism research batches process without hitting size limits.