A PDF is a sealed container. Getting something out of it means choosing between two fundamentally different operations, and most confusion here comes from not realising they are different.
Rendering turns a page into a picture of itself: everything visible survives, and nothing stays as text. Extraction reaches inside for objects that are already there - the characters in the text layer, or the images the author embedded. One preserves appearance, the other preserves meaning, and no tool gives you both.
Which one do you need?
You want the page to look exactly as it does now
PDF to JPG for pages that are mostly photographs or scans. PDF to PNG for text, graphs and line diagrams, where JPG compression puts visible haloes around every sharp edge. Both render the page; neither keeps the text searchable.
You want the words
PDF to Text reads the text layer out as a plain file. It also reports how many pages had no text at all, which is the fastest way to find out whether you are holding a real document or a photograph of one.
You want a figure at full quality
Extract Images pulls embedded pictures out at the resolution the author supplied, which is often far higher than the page displays at. If nothing comes back, the graphics are vector drawings rather than images - render the page instead.
At a glance
| Tool | Keeps appearance | Keeps text | Best for |
|---|---|---|---|
| PDF to JPG | Yes | No | Photographic pages and scans |
| PDF to PNG | Yes, exactly | No | Text, graphs and line diagrams |
| PDF to Text | No | Yes | Searching, quoting, note-making |
| Extract Images | The image only | No | Figures at original resolution |
Before you keep the result
- 150 DPI is right for anything read on a screen. 300 only if it will be printed at full size.
- If text extraction returns nothing, the document is a scan. That is a fact about the file, not a tool failure.
- Once a page becomes an image, the text is gone for good. Keep the original PDF.
- Extracted material is still owned by whoever owns it. Personal revision use and republishing are not the same thing.
Questions and answers
JPG or PNG?
JPG for photographs and scans, where its compression is efficient and its artefacts invisible. PNG for text and line art, where JPG smears faint haloes around sharp edges.
Why can I not get the text out of my paper?
Because it has no text layer. It is a scan - a photograph of a page - and the letters in it are pixels. Only optical character recognition turns those back into characters, and this site does not pretend to do that.
Can I convert a PDF back to Word?
Not here. A faithful PDF-to-Word conversion has to reconstruct a document structure that the PDF never stored, and the results are usually worse than rebuilding the document from extracted text. That is an honest limitation rather than a missing feature.