Two related questions with the same answer: you probably do not need to download anything.
Why you might want a sample file
- Testing an upload form — size limits, accepted types, error handling.
- Testing a printer or a print pipeline.
- Checking a viewer renders correctly after a change.
- Testing OCR against a known page of text.
- Filling a placeholder in a demo or a screenshot.
Each of these wants slightly different properties, which is the argument for making the file rather than finding one — a downloaded sample has whatever characteristics it happens to have.
Make one in a minute
- Open anything — a text editor, a browser page, a word processor.
- Add known content: a heading, a paragraph, a numbered list, a table, an image.
- Print to PDF. Ctrl+P, then Microsoft Print to PDF or Save as PDF.
- For a large file, insert a few full-resolution photographs.
- For a multi-page file, repeat the content or merge the result with itself.
- For a scanned-looking file, print it, photograph it, and convert the image back.
That last one is genuinely useful for testing OCR, because it produces a file with no text layer — which is the condition you actually need to test against, and which a clean generated PDF will never reproduce.
Sizes you may want
| To test | Make |
|---|---|
| Upload limits | Files at 1, 5, 10 and 25 MB |
| Page handling | A 1-page, a 50-page and a 500-page file |
| Rendering | A file with an embedded font and a table |
| OCR | A photographed page with no text layer |
| Compression | A file of full-resolution photographs |
| Merging and splitting | Several small files with numbered pages |
Numbering the pages in the content is a small thing that makes merge and split testing far easier — you can see at a glance whether the output is in the right order, which a file of identical pages will not tell you.
On downloading unknown PDFs
PDFs have been a malware delivery mechanism for a long time. The format supports embedded JavaScript, launch actions and embedded files, and PDF readers have a long history of parsing vulnerabilities.
Downloading arbitrary PDFs from unfamiliar sites in order to test PDF handling is a poor trade, given that generating one takes a minute.
Where you do need to open an unknown PDF, a browser viewer is the safer choice. Chrome and Edge render in a heavily sandboxed process that is updated constantly, whereas desktop readers are frequently years out of date on the machines that have them.
Reading PDFs on Windows
Windows has included a capable PDF viewer in Edge for years, and for reading, searching, filling simple forms and basic annotation it is sufficient. Installing anything is a choice rather than a requirement.
Where something more is needed, the options differ more than the marketing suggests.
| Reader | Good for | Note |
|---|---|---|
| Edge (built in) | Reading, annotation, forms | Already installed, sandboxed |
| Adobe Reader | XFA forms, signature validation | Heavy; the only option for some tasks |
| SumatraPDF | Speed on large files | Open source, minimal, read-only |
| Okular | Annotation, many formats | Open source, cross-platform |
| PDF-XChange Editor | Editing and OCR | Free tier is generous |
Adobe Reader remains necessary for two things and not much else: XFA forms, which some government and banking systems still use, and validating digital signatures properly. If neither applies, Edge or SumatraPDF will serve better and start faster.
A word about Windows 10
Windows 10 reached end of support in October 2025. Machines still running it no longer receive security updates unless they are on an extended support programme.
This matters more for PDF handling than for most tasks, because opening documents from outside the organisation is precisely the activity that exploits a parsing vulnerability.
If you are choosing a PDF reader for a Windows 10 machine, the reader is the smaller question. Prefer one that still receives updates on that platform, open unknown files in a browser viewer rather than a desktop reader, and treat the operating system itself as the item that needs addressing.
Frequently asked questions
Where can I get a sample PDF for testing?
Make one — print any document to PDF and you have a file whose properties you know. That is faster than searching, and it lets you produce the specific size, page count or characteristics your test needs.
Is it risky to download test PDFs?
It is an unnecessary risk. PDFs are a long-established malware delivery format supporting embedded JavaScript and launch actions, and readers have a long history of parsing vulnerabilities. Generating your own takes about a minute.
How do I make a PDF with no text layer for OCR testing?
Print a page, photograph it, and convert the image back to PDF. That produces a genuine image-only file, which is the condition OCR needs to be tested against and which a generated PDF never reproduces.
What is the best free PDF reader for Windows?
Edge, which is already installed and handles reading, search, simple forms and annotation in a sandboxed process. SumatraPDF is faster on very large files. Adobe Reader is needed only for XFA forms and proper digital signature validation.
Do I need to install a PDF reader on Windows?
No. Edge includes a capable viewer and it is kept current automatically, which is also a security advantage over a desktop reader that may go years without an update.
Is it safer to open unknown PDFs in a browser?
Generally yes. Browser viewers run in a heavily sandboxed process and update constantly, while installed desktop readers are frequently out of date on the machines that have them.
Does Windows 10 still get security updates?
No — it reached end of support in October 2025, and machines outside an extended support programme no longer receive them. That matters particularly for opening documents from outside your organisation.