Tiny Online Tools
Why "does this PDF have text?" is the wrong question
A PDF can be searchable on page one and a photograph on page two. The usual check — extract text, find some, declare victory — gets both of those wrong, and it gets the most common real-world case wrong too: a scan that has already been through OCR, which is genuinely searchable even though every visible pixel is an image.
This detector works page by page and gathers six independent signals before it says anything:
- Extracted text — how many words and characters a text extractor actually gets back.
- Text render modes — glyphs drawn in mode 3 are invisible. That is exactly how OCR software writes recognised words over a scan, and it is the only reliable way to tell an OCR'd scan from a born-digital page.
- Drawn images — where each image lands on the page and what fraction of the page it covers.
- Effective DPI and bit depth — a 1-bit image compressed with CCITT or JBIG2 is a scanner's output, not a photograph or a logo.
- Vector paths — a page laid out by a word processor almost always has rules, borders or line art. A bare scan has none.
What you get
Each page is labelled born-digital text, scanned image, scan + OCR layer, mixed, or empty, with a confidence and — this is the part that matters — the list of facts that produced the label. If you disagree with a verdict, you can see precisely which signal drove it.
The document summary answers the question you actually came with: is this searchable end to end, and if not, which pages need OCR. Export the whole report as CSV or JSON for an archiving workflow.
Limits
Nothing inside a PDF declares "I am a scan", so this is inference. A page whose text has been converted to outlines has no text objects and looks identical to a scan. A scan with a poor OCR layer is technically searchable but practically useless. Both are reported with their evidence rather than a confident-sounding guess.
Privacy
The file never leaves your computer. Everything — parsing, rendering, thumbnails, the report — runs in your browser.