Why does text copied from my PDF come out garbled?

A PDF stores glyph numbers, and a reader needs a ToUnicode map (or a standard encoding) to know which character each glyph represents. Many composite fonts with Identity-H encoding and Type3 fonts have no ToUnicode map, so copy and search break. These fonts are flagged; the usual fix is to export the file again from its source, or to add a text layer with OCR.

What does ‘not embedded’ mean for me?

A non-embedded font is referenced only by name, so the reader uses a font with that name from your computer or substitutes a similar one; different widths can change line breaks and layout, which is why print shops require embedded fonts. The standard 14 fonts (Helvetica, Times, Courier and so on) are built into every reader, so leaving them out is fine.

Can I install and use the extracted fonts?

Fonts in PDFs are usually subsets containing only the characters the document uses, and they often lack tables needed for installation, so they don’t work as complete fonts. More importantly, fonts are copyrighted: extracted files are meant for troubleshooting and checking which fonts are used. Make sure you hold a licence before redistributing them or using them in other projects.

Is my file uploaded? Why does a scan show no fonts?

Nothing is uploaded; the file is parsed in your browser. Pages of a scanned or image-only PDF are pictures, so there is no text and therefore no font. Scans that have been OCR’d usually show a single font used for the invisible text layer.

🆎

🆎PDF Font Checker & Extractor - Embedded, Subset, ToUnicode

Need to know which fonts a PDF uses, got a print-shop complaint that fonts aren’t embedded, or wondering why copied text comes out garbled and searches find nothing? This tool lists every font in the file: its name (with ABCDEF+ subset prefixes stripped and shown separately), type (Type1, TrueType, Type0 composite with its CIDFontType0/2 descendant, Type3), whether and how it is embedded, its encoding (WinAnsi, Identity-H, custom Differences and so on), whether it has a ToUnicode map, the pages it appears on and its size — or narrow it down to a single page. Likely problems are flagged automatically: Identity-encoded fonts and Type3 fonts without ToUnicode usually copy as gibberish and can’t be searched; fonts with custom glyph names such as g12 and no ToUnicode are risky too; and fonts that are neither embedded nor among the standard 14 are substituted on computers that don’t have them, which can shift the layout. Embedded font programs can be downloaded one by one or as a ZIP: FontFile2 as .ttf, FontFile3 as .cff or .otf, and Type1 FontFile wrapped with segment headers as .pfb, ready for FontForge and similar tools. TrueType subsets are previewed by drawing their glyph outlines directly; OpenType fonts are loaded by the browser where possible to show their mapped characters. Everything runs in your browser and nothing is uploaded. Remember that fonts are copyrighted software — extraction is for inspection and troubleshooting.

Free · No sign-up · Runs in your browserLast updated: 2026-10-26

Loading tool...

How to use PDF Fonts

1Upload a PDF to see every font’s type, embedding, encoding, ToUnicode and pages
2Switch to a single page if needed and review the flagged copy, search and embedding problems
3Preview glyphs, download a single embedded font, or get all of them as a ZIP

FAQ

Why does text copied from my PDF come out garbled?: A PDF stores glyph numbers, and a reader needs a ToUnicode map (or a standard encoding) to know which character each glyph represents. Many composite fonts with Identity-H encoding and Type3 fonts have no ToUnicode map, so copy and search break. These fonts are flagged; the usual fix is to export the file again from its source, or to add a text layer with OCR.
What does ‘not embedded’ mean for me?: A non-embedded font is referenced only by name, so the reader uses a font with that name from your computer or substitutes a similar one; different widths can change line breaks and layout, which is why print shops require embedded fonts. The standard 14 fonts (Helvetica, Times, Courier and so on) are built into every reader, so leaving them out is fine.
Can I install and use the extracted fonts?: Fonts in PDFs are usually subsets containing only the characters the document uses, and they often lack tables needed for installation, so they don’t work as complete fonts. More importantly, fonts are copyrighted: extracted files are meant for troubleshooting and checking which fonts are used. Make sure you hold a licence before redistributing them or using them in other projects.
Is my file uploaded? Why does a scan show no fonts?: Nothing is uploaded; the file is parsed in your browser. Pages of a scanned or image-only PDF are pictures, so there is no text and therefore no font. Scans that have been OCR’d usually show a single font used for the invisible text layer.