format

EPUB

EPUB is the open ebook standard, a ZIP of HTML that reflows to any screen.

Convert a file now

What is a EPUB file?

An open ebook format packaging XHTML, CSS and images in a ZIP, with text that reflows to the reader's screen and font size.

Technical characteristics

Representation
binary
Compression
Deflate
Lossy
No
Open specification
Yes
Encryption support
Yes

History and origin

Introduced
2007
Organisation
IDPF, now W3C

Compatibility

  • Windowspartial
  • macOSfull
  • Androidfull
  • iOSfull

Applications: Apple Books, Kobo, Calibre, Google Play Books.

Advantages of EPUB

  • Reflows to any screen size
  • Adjustable fonts and accessibility support
  • Open standard

Limitations to know about

  • Not supported by Kindle's older formats
  • Fixed layouts (comics, textbooks) render inconsistently

When to use EPUB

  • Novels and long-form reading
  • Library lending
  • Self-publishing

Convert EPUB files

EPUB to PDF fixes the layout and loses reflow, which is a downgrade for e-reader use.

EPUB scores

Deterministic scores calculated from the structured facts above. Identical facts always produce identical scores.

Overall score

81/100

  • Editing score

    100
    • lossless data survives repeated open-edit-save cycles without degrading
    • a broad range of applications read and write it

    Confidence 51% · evidence coverage 63%

  • Longevity score

    100
    • it is actively maintained and current
    • the specification is published openly
    • the specification is public, so a decoder can always be rebuilt
    • it sits at the centre of the format ecosystem, with many documented relationships

    Confidence 82% · evidence coverage 100%

  • Archiving score

    96
    • the specification is public, so a decoder can always be rebuilt
    • the specification is published openly
    • it is actively maintained and current
    • an openly specified format can be implemented without a licence negotiation

    Confidence 83% · evidence coverage 89%

  • Compatibility score

    96
    • Windows, macOS, Linux, Android and iOS all open it without extra software
    • a broad range of applications read and write it
    • an open specification means cloud SDKs can implement it without licensing friction
    • it opens by double-clicking on every mainstream device

    Confidence 72% · evidence coverage 75%

  • AI score

    92
    • open specifications are implemented by the open-source decoders that ML stacks rely on
    • the specification is published openly
    • a broad range of applications read and write it
    • large lossless payloads cost bandwidth and memory

    Confidence 67% · evidence coverage 100%

  • Openness score

    87
    • the specification is published openly
    • an openly specified format can be implemented without a licence negotiation

    Confidence 84% · evidence coverage 100%

  • Printing score

    78
    • lossless data means no compression artefacts appear at print resolution
    • a broad range of applications read and write it

    Confidence 59% · evidence coverage 50%

  • Photography score

    75
    • lossless data survives repeated open-edit-save cycles without degrading
    • it opens by double-clicking on every mainstream device
    • it stores data losslessly, so files stay large

    Confidence 61% · evidence coverage 44%

  • Web score

    41
    • the specification is published openly
    • it stores data losslessly, so files stay large
    • large lossless payloads cost bandwidth and memory

    Confidence 66% · evidence coverage 67%

  • Performance score

    35
    • it stores data losslessly, so files stay large
    • large lossless payloads cost bandwidth and memory

    Confidence 67% · evidence coverage 100%

  • Streaming score

    33
    • it stores data losslessly, so files stay large
    • large lossless payloads cost bandwidth and memory

    Confidence 70% · evidence coverage 50%

Every conclusion below is derived from the structured facts on this page — nothing is written by hand.

  • Operating system compatibility100/100

    • Because Windows, macOS, Linux, Android and iOS all open it without extra software.
    • Because most platforms handle it natively.
  • Software compatibility100/100

    • Because a broad range of applications read and write it.
    • Because several mainstream applications support it.
  • Cloud compatibility100/100

    • Because an open specification means cloud SDKs can implement it without licensing friction.
    • Because it converts into many other formats, which suits automated pipelines.
  • AI compatibility100/100

    • Because open specifications are implemented by the open-source decoders that ML stacks rely on.
    • Because lossless pixels avoid compression artefacts leaking into training data.
    • Because wide library support means common data loaders can read it.
  • Long-term archiving100/100

    • Because the specification is public, so a decoder can always be rebuilt.
    • Because lossless storage preserves the original data for future migrations.
    • Because it is a current, actively maintained format.
  • Professional editing100/100

    • Because lossless data survives repeated open-edit-save cycles without degrading.
    • Because professional applications read and write it.
  • Open standards100/100

    • Because the specification is published openly.
    • Because an official specification document is publicly reachable.
  • Popularity100/100

    • Because it sits at the centre of the format ecosystem, with many documented relationships.
    • Because mainstream applications ship support for it.
    • Because every mainstream operating system handles it.
  • Future-proofing100/100

    • Because it is actively maintained and current.
    • Because open specifications outlive the companies that create them.
    • Because nothing in the graph supersedes it yet.
  • Everyday consumer usage75/100

    • Because it opens by double-clicking on every mainstream device.
    • Because it is a current format that people and services still expect.
  • Printing suitability71/100

    • Because lossless data means no compression artefacts appear at print resolution.
  • Compression efficiency23/100

    • Because it stores data losslessly, so files stay large.
    • Because its codec generation predates modern entropy coding and block partitioning.
  • Web suitability33/100

    • Because lossless payloads cost more bandwidth than the web usually justifies.

File extensions

Compared with

Supported by

Created by

Specified by

Compression used

Alternatives

Frequently asked questions

Can I read EPUB on a Kindle?

Yes — modern Kindles and Send to Kindle accept EPUB and convert it internally. Older devices need MOBI or AZW3.

Data quality

Every claim on this page is classified as a verified fact, a derived fact or an assumption.

Overall confidence
88%
Verification status
verified
Last reviewed
2026-08-03
Verified facts
3

3 classified claims backed by 1 source.

Sources and review status

Review status: verified · confidence 95%

Last reviewed 2026-08-03 (revision 1).