Zopfli Library Improves Zlib Compression by 3 to 8%

Google developers have released a new compression library called Zopfli. This library, written in C, is compatible with zlib, yet provide a better compression, more exactly 3 to 8% according to Google.

This library can be used on servers for better compression in order to save bandwidth, as well as delivering web pages faster. Since it’s fully compatible with zlib, the web browsers do not need to be changed. The only drawback is that it’s several magnitude slower than zlib, so it’s better used for static content that is compressed once, and sent over the Internet many times, and it may not be a good choice for dynamic content.

The source code is available at https://code.google.com/p/zopfli/, so let’s try it.

Get the code and build zopfli:

Different levels of compression are available:

For testing purpose, I’ve just saved this blog as one html file (test.html – 67275 bytes) with different options to see the compression improvements and time it takes with zopfli.

Command line used for the test:

Here are the results (The first line is the compression with gzip (gzip -9 test.html):

OptionFile Size (Bytes)ImprovementTime (s)
Gzip -9153370.011

If you run zopfli without options it will output to gzip format with 15 iterations (i15). With this particular html file, zopfli can achieve around 2.6% better compression than gzip -9, but it’s way slower on my machine, with the fastest option being about 17x slower.

Running gzip -d test.html.gz on a file compressed with zopfli, just works as advertised.

Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.

Support CNX Software - Donate via PayPal, become a Patron on Patreon, or buy review samples
Notify of
newest most voted
7 years ago

But how does it compare to other free compression bzip2, or 7zip-lzma?

7 years ago

They’d be better off just supporting a more modern compressor in the client, they could easily get mozilla on-board which would stick up the majority of the web.

The improvement just from the major segment of the browser market would probably more in total than just this modest improvement.

e.g. lzma gets nearly 9% on the front page, at about 450ms (unusually high sys time though)