Mac OCR Guide

How to copy text from a PDF on Mac when selection fails

Some PDFs behave like normal documents. Others are effectively stacks of images. When text selection fails in Preview or another viewer, GlassCopy gives you a fast fallback that works on the visible page.

See GlassCopy for Mac Browse all guides

Soft holographic composition of a scanned PDF page with one clause isolated inside a glass OCR capture frame. — Flattened PDFs are easier to handle when you capture one clause, field, or table section at a time.

Why copy fails here

A PDF may contain an embedded image instead of live text, or a complex export may make the visible layer hard to copy cleanly even though it looks like a document.

A scanned contract opens in Preview, but

A scanned contract opens in Preview, but dragging across the paragraph never highlights any words.

An invoice PDF exports from another system

An invoice PDF exports from another system as a picture, so account numbers and totals are not selectable.

A presentation deck saved as PDF contains

A presentation deck saved as PDF contains flattened slides with readable text but no live text layer.

Try the cleanest source first

Native copy check

Try selecting text in Preview or your PDF reader first. Use GlassCopy when the page is a scan, a flattened export, or a locked visual layer where dragging never highlights words.

Why GlassCopy helps

GlassCopy works on what is visible, so it is useful when the PDF file does not expose a reliable text layer but the page is readable on screen.

How to do it with GlassCopy

GlassCopy is useful when a PDF page looks readable but the PDF viewer cannot highlight or copy the text because the content is image-based.

Check whether the PDF page is image-based

Try selecting a few words in Preview first. If you only get a whole-page selection or nothing at all, treat the page as an OCR job.

Zoom the page until the target text is crisp

Increase zoom before capturing, especially for small footnotes, invoice fields, or dense slide exports.

Use GlassCopy on one region at a time

Capture a column, table block, or paragraph rather than the whole PDF page so the OCR engine reads consistent text structure.

Paste into a plain-text field and clean formatting manually

Review line breaks and table spacing after paste, because PDFs often introduce layout-driven breaks that are different from the semantic content.

What to capture, what to leave out

Select

One clause, paragraph, table column, invoice field, or slide-export section at a time.

Leave Out

Headers, footers, signatures, page numbers, watermarks, and neighboring columns unless they change the meaning.

Best Conditions

Use the PDF viewer zoom so the target text is crisp, then capture a tight rectangle around the content.

Tips that improve results

Helpful habits

For tables, capture one column or section at a time so numbers stay aligned.
Use the PDF viewer's zoom instead of macOS display scaling so the text edges are sharper.
If you need several fields from a multipage PDF, work page by page rather than one giant OCR pass.

Avoid these mistakes

Do not assume every non-copyable PDF is encrypted; many are simply scanned images.
Avoid including page headers and footers when you only need body text or totals.

When not to use this workflow

If the PDF already has a searchable text layer and copy works cleanly, use the native selection tool instead.

Clean up and verify

After OCR

Rejoin line breaks caused by narrow PDF columns.
Verify numbers, totals, section references, and legal citations against the visible source.
For tables, paste into a plain-text scratch area before moving values into a spreadsheet.

Accuracy watchlist

Scanned PDFs can confuse punctuation, footnote markers, and small superscripts.
Two-column layouts may paste in the wrong reading order if captured too broadly.
Watermarks and stamps can be mistaken for body text when they overlap the selection.

Privacy boundary

For contracts, invoices, and IDs, capture the smallest useful region and redact names or account numbers before sending text to another service.

Real situations where this guide helps

Pulling clauses from a signed scanned contract

Pulling clauses from a signed scanned contract into an email summary.

Copying totals and invoice numbers from accounting

Copying totals and invoice numbers from accounting PDFs into a spreadsheet.

Extracting bullet points from a slide deck

Extracting bullet points from a slide deck exported as PDF.

Example workflow

Source

A scanned invoice PDF shows the vendor, invoice number, and total but the page will not select.

Selection

Select the invoice number and total as two small regions rather than the full page.

Result

Accurate fields ready for a note, spreadsheet, or accounting system.

Destination

Spreadsheet row, reimbursement form, or client email.

Questions people ask

Why can I see PDF text but not select it?

Many PDFs are scans or flattened exports. The words are visible pixels, but the file does not expose a reliable text layer.

When should I use Preview instead of GlassCopy?

Use Preview or your PDF reader when normal selection already works. Use GlassCopy when dragging over the page never highlights the text you need.

What is the best way to OCR PDF tables?

Capture one field, column, or small table section at a time, then verify numbers before using them downstream.

Related guides

These pages cover adjacent workflows without repeating the same advice.

How to copy text from scanned documents on Mac

Scans often have uneven lighting, skew, stamps, and low contrast. The right workflow is to isolate the useful text region instead of treating the whole page as one OCR block.

How to copy text from slides on Mac

Slide OCR is easiest when you isolate a title, bullet stack, or chart caption rather than capturing the entire visual slide composition.

What to do when text is not selectable on Mac

The fastest way to solve non-selectable text is to identify whether you are looking at an image, a remote session, a flattened document, or a protected app surface, then choose the right workaround.