How to Extract Text from a PDF or Image Using OCR (Free & Online)
March 2026 · 5 min read
You've got a scanned PDF, a photo of a document, or an image with text, and you need the actual words out of it. Copy-paste doesn't work. Typing it out manually is out of the question. This is exactly what OCR (Optical Character Recognition) is for, and you can do it free in seconds without installing anything.
What Is OCR?
OCR stands for Optical Character Recognition. It's a technology that analyzes an image, whether it's a scanned document, a photograph, or a PDF created from a scan, and converts the visual text in that image into actual, selectable, copyable characters.
Without OCR, a scanned PDF is just a picture. With OCR, it becomes a real document you can search, copy, translate, or paste into another file.
When Do You Need OCR?
You'll need OCR whenever you have text that's "trapped" inside an image:
- A scanned contract or invoice saved as PDF
- A photo of a whiteboard or handwritten notes
- An old document that was digitized by scanning
- A screenshot containing text you want to copy
- A PDF created from a scanner where text is not selectable
- A business card photo you want to extract contact details from
How to Extract Text Using Convrex OCR, Step by Step
Here's how to use the Convrex OCR Tool, no account, no software, no cost:
Select a PDF or image file (JPG, PNG, etc.) from your device. Files up to 100MB are supported.
The tool first tries to extract embedded text directly from the PDF. If the document is a scan or image-only, it automatically runs OCR to recognize the text visually.
The extracted text appears on screen. Copy it directly, or download it as a plain text file.
No sign-up. No watermarks. Files deleted immediately after processing.
Extract Text Free
What's the Difference Between a Text PDF and a Scanned PDF?
A text PDF was created digitally, exported from Word, generated by software, or printed to PDF from a browser. The text in it is real data: you can click, select, and copy it.
A scanned PDF was created by physically scanning a paper document. The result is essentially a photo embedded in a PDF container. There's no selectable text, just pixels arranged to look like letters. OCR is what bridges the gap.
Convrex handles both automatically: it tries direct text extraction first, and falls back to OCR if the document is image-only.
Tips for the Best OCR Results
Use clear, high-contrast scans. OCR accuracy drops significantly on low-resolution or blurry images. A 300 DPI scan in black and white gives near-perfect results.
- Avoid skewed pages, a document scanned at an angle confuses OCR engines. Scan flat or straighten the image first.
- Higher resolution = better accuracy, 300 DPI is the sweet spot. Below 150 DPI, error rates increase sharply.
- Plain fonts work best, OCR handles Arial, Times New Roman, and similar fonts with near-perfect accuracy. Decorative or handwritten fonts are harder.
- Clean backgrounds help, coffee stains, shadows, or noise on a page will appear as extra characters in the output.
- Handwriting is hit or miss, printed text is reliably recognized; cursive handwriting is much harder for OCR and results will vary.
What Happens to My File After Processing?
Your file is processed entirely on Convrex servers and permanently deleted the moment your result is delivered. Nothing is stored, cached, or retained. You can verify this in our Privacy Policy.
Related Tools You Might Need
- PDF to Text, extract text from a native (non-scanned) PDF directly
- PDF to Word, convert your PDF into an editable .docx file
- Compress PDF, reduce file size before or after OCR processing
- Image to PDF, convert your image to PDF first, then run OCR on it
Summary
OCR turns a static image of text into real, usable characters. Whether you're working with a scanned contract, a photographed receipt, or an old archived document, you can extract the text in seconds, for free, directly in your browser. Upload your file, hit extract, and you're done.
Try OCR Free
Frequently Asked Questions
Does OCR work on all languages?
Convrex OCR uses Tesseract, which supports over 100 languages. For best results with non-Latin scripts (Arabic, Chinese, Japanese, etc.), make sure the scan quality is high.
Can OCR read handwriting?
Printed text: yes, with high accuracy. Cursive or stylized handwriting: results will vary. OCR is optimized for printed characters, not handwriting recognition.
My PDF already has text but it's still wrong, why?
Some PDFs embed text but the content is garbled or incorrectly encoded (common with older scanned PDFs that had OCR run on them badly). In that case, re-running OCR on the page images can sometimes produce cleaner output.
Is there a page limit?
No page limit, just a 100MB file size limit, which covers all typical use cases including multi-page scanned documents.
Can I OCR a photo taken on my phone?
Yes. Upload the photo (JPG or PNG) directly. Make sure the image is in focus and the text fills most of the frame for the best accuracy.