How to search inside many PDFs at once

Last reviewed: September 23, 2026 · Markdown version

The blunt version: a reader's find function searches the file that is open, so what you're asking for — type a word once, check every document you own — is a different job. There are two honest routes, and which one you need depends on how many documents you have and how often you hunt through them.

Route 1: one at a time, by hand

Open a file, use its find, close it, repeat. For three known files this is the right answer — no app needed. It breaks down in two places:

Volume with uncertainty. When you don't know which of forty statements, contracts, or papers holds the clause you remember, one-at-a-time means opening all forty — and doing it again for the next question.
Scans return nothing. A scanned page is a picture of text — invisible to ordinary search. An in-document search of a scan quietly finds zero, even when the word is printed right there. Why, and how to tell.

Route 2: index the library once, then search everything

The structural fix is an index: a searchable table of every word in every document, built once up front. After that, one query checks the entire library and every later search reads the index instead of re-reading the documents. SQLite's documentation for its full-text module puts the goal in one line: such engines "allow the user to efficiently search a large collection of documents for the subset that contain one or more instances of a search term" (SQLite FTS5 Extension, read 23 September 2026). The catch on a phone is that something has to build and maintain that index — and handle the scans, which need OCR before they contain any searchable words at all.

Where DocFind fits

DocFind is route 2, on your phone, honestly labelled. It builds a searchable index of the PDFs you add — scanned ones too, read on your device (Apple's Vision on iPhone, Google's ML Kit bundled into the app on Android). One search covers the library, and each result shows the file, the page and the text around the match; tap a result and that page opens with your keywords highlighted. Matches that came from OCR are labelled, and pages that couldn't be made searchable are reported — so an empty result tells you the word isn't in what was read, and you know what wasn't. Your documents never leave the device: searching, indexing and reading scanned pages all work with no internet connection.