How to Recognize Text in PDF: 6 Ways to Turn a Scan Into Selectable Text

Why your PDF won't let you copy: you tried to select a sentence and nothing highlighted, Ctrl+F finds no matches, and pasting into a translator gives you nothing — so you're stuck retyping by hand. That's because the file is a scan or photo: every page is just an image, with no real text underneath. OCR (Optical Character Recognition) reads that image and rebuilds a genuine text layer, turning the picture back into words you can select, search, copy, and edit. The six methods below all do this — the only choice is whether you want the whole file recognized, just one region, or a quick browser fix.

Which Method Do You Need? (Quick Routing Table)

Match your situation to the right method, then jump to that section. UPDF covers every row — on desktop, mobile, and in the browser.

Part 1. How to Recognize Text in a Scanned PDF with Desktop OCR

When you open a scanned PDF and nothing selects, UPDF's OCR tool reads the page image and rebuilds a genuine text layer, so the once-flat scan behaves like a normal PDF — you can select, search, copy, and edit. UPDF's OCR tool recognizes text in 38 languages and preserves the original layout and fonts.

Open the document in UPDF — if you don't have it yet, download UPDF for free and install it in under a minute — then follow the steps.

Step 1. Launch UPDF and click Open File to import your scanned PDF. For a loose image, drag it onto the UPDF home window to open it.

Step 2. Go to leftside Tools → OCR.

Step 3. Under Document Type, choose Editable PDF (text over or under the page image) so you can both search and edit. Pick Searchable PDF Only instead if you only need to find text without changing it.

Step 4. Set the document language and page range. You can tick more than one language at once for a mixed-language file — no need to split it.

Step 5. Click the gear icon next to the Layout dropdown for advanced options such as MRC compression, then click Convert.

Step 6. When OCR finishes, the recognized PDF opens in a new tab. Click Edit in the top toolbar to correct text, or just select and copy what you need.

Accuracy reminder: OCR is recognition, not magic — always proofread numbers in invoices, IDs, and contracts, since a misread "0/O" or "1/l" changes meaning. A clean 300 DPI scan reads far more accurately than a dim, skewed phone photo.

Part 2. How to Recognize Text in a Scanned PDF on iPhone, iPad & Android

One UPDF license covers desktop and mobile, so you can recognize text on the device in your hand. UPDF's mobile OCR is full-featured — the same three output modes as desktop (Editable PDF, Text and Pictures Only, Searchable PDF Only) — not a stripped-down version. Install UPDF for iOS or UPDF for Android, then open your file. The steps are identical on both platforms.

Step 1. Tap the + button to import your scanned PDF (or an image).

Step 2. Tap OCR at the bottom, set the language and layout, then tap Continue. The recognized file saves to your UPDF files automatically.

Step 3. Tap Edit on the OCR'd file to select, copy, or rewrite the text, then save.

Part 3. How to Recognize Handwriting or a Single Passage with Desktop AI

When OCR struggles — messy handwriting, a stylized font, a chart you can't select — UPDF's built-in AI reads the image directly and returns the text. On desktop you don't even need to export anything: the Screenshot tool grabs the region straight off the open PDF and sends it to the AI. This is the chat with image feature, and it handles handwritten and printed text in effectively any language.

Step 1. Open the PDF in UPDF, then click UPDF AI in the lower-right corner and switch to Chat mode at the top.

Step 2. Navigate to the page you need, then click the Screenshot tool at the bottom right.

Step 3. Click and drag over the area with the text or handwriting. When you release, the screenshot lands in the UPDF AI message box.

Step 4. Type a prompt — Extract the text from this image — and hit Send. UPDF AI returns the text in seconds; click the Copy icon to paste it wherever you need.

Part 4. How to Recognize Text From an Image on Your Phone with AI

On mobile, UPDF AI's chat-with-image works from an uploaded picture rather than an in-app screenshot — so snap or save the image first, then hand it to the AI. This is the practical route when the text lives in a photo: a whiteboard shot, a printed note, a foreign menu.

Step 1. Open UPDF on your phone and tap UPDF AI, then switch to Chat mode.

Step 2. Tap the image icon and select the photo containing the text.

Step 3. Enter a prompt such as Extract text from this image and send. Copy the recognized text and paste it into any app.

Part 5. How to Grab Text From Any Area with Right-Drag "Copy as Text"

UPDF has a faster shortcut for a single region: hold the right mouse button and drag a box around the area you want. When you release, a floating menu appears with Copy as Image and Copy as Text — choosing Copy as Text recognizes the text inside that box and puts it straight on your clipboard. There's no full-file OCR pass and no AI prompt; you box the region and paste.

Step 1. Open the PDF in UPDF. Right-click and drag to draw a box around the paragraph, table cell, or labeled area you want.

Step 2. Release the mouse. In the floating menu, click Copy as Text (or Copy as Image if you'd rather keep the region as a picture).

Step 3. Paste the recognized text into Word, an email, or any editor with Ctrl + V / Cmd + V.

Part 6. How to Recognize Text in a PDF Online (No Install)

On a borrowed or locked-down computer where you can't install software, UPDF AI Online recognizes text right in the browser. It gives you two paths, depending on whether you want the whole file converted or just one chunk of text.

Path A — Auto-OCR a whole scanned PDF:

Step 1. Click the button below to open UPDF AI Online in your browser.

Step 2. Upload your scanned PDF and click Send.

Step 3. UPDF AI Online runs OCR automatically and then you can copy the text.

Path B — Extract text from one image or screenshot:

Step 1. In UPDF AI Online, click the image icon (or paste a screenshot straight into the message box).

Step 2. Upload a JPG or PNG of the page or region — handwritten or printed.

Step 3. Enter Extract the text from this image, send, and copy the result.

Privacy and limits: the online route uploads your file to UPDF's servers, so for contracts, IDs, and financial documents the desktop app (which processes everything locally) is the safer choice. Online free use covers 5 documents, and it's best for short files — for large or 100-page-plus PDFs, use the desktop OCR in Part 1.

Download UPDF for free to try recognizing text on a scan of your own — installation is free, and Pro features are available when you need watermark-free export or batch OCR.

Part 7. Comparison: Which UPDF Method Fits Your File?

Every method recognizes text, but each has a boundary worth knowing before you start.

Mac note: UPDF's desktop OCR runs on the web-download version for Apple Silicon Macs and on Windows; the Intel-Mac and Mac App Store builds don't include OCR. On those, use the screenshot route or UPDF AI Online instead.

Part 8. Edge Cases: When Recognition Still Fails

• You ran OCR but the text still won't select. The scan is likely too low-resolution or skewed. Re-scan at 300 DPI, straighten (deskew) the page, and run OCR again — recognition quality depends heavily on a clean, upright image.
• The recognized text is full of wrong characters. You probably left the wrong document language set. An English-only pass on a French, German, or Arabic file misreads accented and non-Latin characters. Re-run with the correct language (or tick several for a mixed-language file).
• OCR is greyed out or blocked. The PDF is permission-protected against editing. Enter the password or remove the restriction first — only on files you own or are authorized to edit — then run OCR. See how to remove PDF security.
• The file is over 300 pages. UPDF's OCR handles up to 300 pages per file. For a longer document, split it, OCR each part, then merge the results back together.
• Handwriting comes back garbled from OCR. Traditional OCR isn't built for cursive. Switch to the AI screenshot route (Part 3) — it reads handwriting far more reliably than a standard OCR pass.

Part 8. FAQ

Conclusion

Recognizing text in a PDF comes down to one question: is it a scan or a digital file? If the text won't select, OCR rebuilds the layer for the whole document; if you only need one passage or some handwriting, UPDF AI's chat-with-image reads it on the spot; and for a single region, right-dragging Copy as Text grabs it in one action. UPDF handles all of these — desktop and mobile OCR, AI screenshot and image recognition, and browser-based UPDF AI Online. If you're recognizing a scanned PDF so you can edit it afterward, the same workflow also lets you make a PDF editable in the same app.

Download UPDF for free to try recognizing text on your own scanned file — installation is free, and Pro features are available when you need watermark-free export or batch OCR.