If you've ever tried to select text in a PDF and nothing happened, or noticed that one PDF is 50 KB while another of the same page count is 5 MB, you've encountered the fundamental divide in the PDF world: text-based PDFs vs. scanned image PDFs. These aren't just different in size — they're different formats underneath, and that difference determines how you can search, edit, compress, and work with them.
A text-based PDF — sometimes called a native PDF — was created from a digital source: a Word document, a web page, a LaTeX file, or design software. Under the hood, the PDF stores the actual text characters as objects, along with font definitions, layout instructions, and any vector graphics or images.
When you open a text-based PDF, the text is selectable. You can copy and paste it. A search for "budget" finds every instance. The file is typically small because text objects are just a few bytes, and any images have likely been compressed already by the originating application.
Common sources: Microsoft Word "Save as PDF," Google Docs download as PDF, LaTeX compiled to PDF, HTML pages printed to PDF, Adobe InDesign exports.
A scanned PDF is created by scanning a physical document — a paper contract, a book page, a form — with a flatbed scanner or a phone camera. The result is a raster image (a grid of pixels) wrapped in a PDF container. There are no text characters stored, no selectable text, no searchable content — just a picture of text.
Because the content is pixel data, these files are typically large — 300 DPI scans of letter-size pages can be 150–300 KB each, and complex documents easily reach 50–100 MB. You cannot copy text from a scanned PDF. You cannot search it with Ctrl+F. If you want the text, you need OCR.
Common sources: Scanner output, fax-to-PDF services, phone scanning apps (CamScanner, Adobe Scan), copier "scan to email."
There are three quick tests:
OCR (Optical Character Recognition) is the process of analyzing a scanned image and recognizing the shapes in it as letters, words, and sentences. When OCR is applied to a scanned PDF, the software creates a hidden text layer and overlays it on the image. The result looks identical visually, but now:
OCR is not perfect — low-quality scans, unusual fonts, or handwriting can reduce accuracy. But for clean, printed documents at 300 DPI or above, modern OCR achieves 98%+ accuracy.
Want to learn more? Read our full guide: What Is OCR? How Optical Character Recognition Works.
| Document type | Page size | Typical file size | Text selectable? | Searchable? |
|---|---|---|---|---|
| Text-based PDF | Letter (8.5"×11") | 10–50 KB per page | Yes | Yes |
| Scanned PDF (300 DPI) | Letter (8.5"×11") | 150–300 KB per page | No | No (without OCR) |
| Scanned PDF with OCR | Letter (8.5"×11") | 150–300 KB per page | Yes (hidden layer) | Yes |
This is where it gets practical. Your compression approach should match the PDF type:
Compression gains are smaller because the file is already small and mostly text. Focus on:
Expect 10–30% reduction. Going beyond that risks losing quality or functionality.
These have the biggest room for compression. You can:
Expect 50–90% reduction. But you'll still have a non-searchable file — consider adding OCR.
Same compression options apply, but keep the OCR text layer intact. If you downsample aggressively, the text layer remains — it's vector text — but verify searchability afterward.
Use text-based PDFs when: You need searchable text, the document will be edited or referenced digitally, you're distributing electronically, or you need small file sizes.
Use scanned PDFs when: You're archiving paper documents as-is, the original formatting must be preserved pixel-for-pixel, or you're submitting to a system that requires an image-based document.
Use scanned PDFs with OCR when: You have paper documents that need to be searchable but don't exist in digital text form.
Our Compress PDF tool detects the PDF structure and applies the right approach:
All processing happens in your browser with WebAssembly — no file uploads.
Compress your PDF with privacy guaranteed
Compress PDF Now (Free)