Search Text in a PDF

When to use Search PDF

Use this to quickly find a clause, name, or figure inside a long PDF — a contract, a report, a manual — without opening it in a separate reader. If the document is a scan with no selectable text, run OCR first (see below).

How to Search Text in a PDF

1. Select a PDF.
2. Type the word or phrase you're looking for.
3. Step through the matching pages using next/previous.
4. Jump to any result to view it in context.

What this tool changes

• Nothing — search only reads the document to build a temporary, local text index.

What it preserves

• The entire original document, unmodified.

Good to know

• Search only works on PDFs with actual selectable text. Scanned documents need OCR first.
• Text extraction order from the PDF can occasionally be imperfect, especially with complex layouts.

It searches the whole document at once

Many readers — particularly in-browser and mobile previewers — search only the page you are looking at, which is why a word you know is in a document sometimes returns nothing. This reads every page and reports each match with its page number and the text surrounding it.

That context line is the useful part on a long document. A list of forty page numbers tells you little; forty snippets tell you which match is the one you want before you go looking.

Everything runs in your browser, so a long confidential document is searched without being uploaded anywhere.

If nothing is found, check what kind of page it is

Try selecting and copying representative text, but treat this as a clue rather than proof. An image-only scan needs OCR; broken character mappings or outlined text can also defeat search.

If text selects normally and search still fails, the likely causes are subtler. Ligatures combine fi and fl into single characters, so "find" may not match. Hyphenation splits words across lines. Curly and straight apostrophes are different characters. Some documents carry a broken font encoding that displays correctly and extracts as gibberish.

The diagnostic is to search a short fragment rather than a whole word — "ocume" instead of "document". If the fragment matches and the word does not, your search term is the problem rather than the file. Why a PDF is not searchable works through all four causes.

Using search to find document boundaries

On a batch scan — twenty invoices captured as one file — search is how you avoid reading forty pages to find where each document starts.

Almost every document type repeats a phrase on its first page: "Invoice number", "Statement period", "Dear", a letterhead address. Search for it and the list of pages it appears on is your list of split boundaries.

Feed those into Split PDF as ranges and a job that looked like an afternoon becomes a few minutes. The prerequisite is a text layer, so on a fresh scan the order is recognise first, then search, then split — set out in how to scan and file paper records.

Searching does not require an export

Searching reads source text without changing the original file. The shared workspace also has an Export control, but that creates a rebuilt PDF; it is unnecessary for searching and should not be used as a preservation step.

No results can mean an image-only scan, but encoding, outlined text, and the exact search term can also explain it. Check the source and compare another capable reader before deciding whether OCR is appropriate.

That is the single most common reason a PDF search fails, and it is not a fault in the search. A scanner, a copier, or a phone camera produces pages with no characters in them at all. Recognising the text is what closes the gap, and why a PDF is not searchable works through the other causes, including the hybrid file where page one searches perfectly and page two finds nothing.