Measured · same site on both stacks · no page cache

Faster than a tuned WordPress, even with the plugins on.

On real WooCommerce and Elementor pages, BookPress renders each page faster than a WordPress tuned with PHP-FPM, OPcache and its JIT, a Redis object cache and MariaDB, at about a tenth of the memory. A site with no plugins renders about 220 times faster. The stacks and every number →

Try the live demo How it works

Time to render a real store page

median time, one request at a time, no page cache · lower is better

BookPress 0.2.40, measured 27 September 2026. WordPress runs its own PHP with OPcache, its JIT and a Redis object cache, with the page cache off.

1.31–1.46×

faster than tuned WordPress on real WooCommerce and Elementor pages, on the Intel server and the ARM server

~220×

faster on a site with no plugins: 0.18 ms against about 42 ms

~10×

less memory with the plugins on: about 170 MB against about 1.7 GB

page caches, object caches or database servers in the request path

The easy case

A site with no plugins: about 220× faster.

A site with no plugins renders with no PHP interpreter in the request path. On the Twenty Twenty-Five theme, the front page renders in 0.18 ms against WordPress's 41.8 ms, in 16.5 MB against 1.07 GB. Brochure sites, blogs and portfolios are this case.

Front page, site with no plugins

Twenty Twenty-Five with the sample content a fresh WordPress ships with, median time, no page cache · lower is better

Twenty Twenty-Five with WordPress's own sample content, on the 4-vCPU Intel server. A blog post is 0.30 ms against 47.4 ms. The condition that travels with this number: no plugins installed at all.

The hard case

The same setup, one page at a time.

The store numbers come from a repeatable setup: it builds a tuned WordPress, loads the identical demo content into WordPress and BookPress, checks that both produce the same HTML, then times one request at a time. BookPress renders every page fresh; the WordPress side runs its own PHP with the page cache switched off, which is the moment a cache steps aside anyway: a cart, a search, an account page, anything with a session.

What both engines served

The real BookPress demo, migrated across: the Kadence theme, WooCommerce with a five-product catalogue, a built Elementor showcase page, Yoast SEO and Contact Form 7. Both stacks produced identical HTML on both servers: identical Elementor markup, identical product page, identical shop catalogue.

The WordPress it beat

PHP 8.3 with OPcache and the JIT enabled, a Redis object cache over a socket, MariaDB, and PHP-FPM with 64 static workers behind nginx. Tuned the way a competent host tunes one. The page cache is off for every measurement, and we check it is off before timing.

Both servers

A 4-vCPU Intel server and a 4-vCPU ARM server (Google Cloud c2-standard-4 and c4a-standard-4), both 16 GB. BookPress wins on both. The ARM server is the faster buy: it renders these pages about 1.3 to 1.5× faster, for both engines.

Every number, as a table

Median time per page, BookPress rendering from scratch against WordPress running PHP with no page cache. BookPress 0.2.40, measured 27 September 2026. Lower is better; the ratio is WordPress ÷ BookPress.
Page	Server	BookPress	WordPress	Faster by
Elementor page	Intel	74.0 ms	97.1 ms	1.31×
WooCommerce shop	Intel	69.7 ms	95.7 ms	1.37×
Elementor page	ARM	49.8 ms	72.9 ms	1.46×
WooCommerce shop	ARM	46.9 ms	68.6 ms	1.46×

BookPress uses about 170 MB of memory on these pages, against about 1.7 GB for the WordPress stack under sustained load, about ten times lighter. Theme and plugin PHP runs on BookPress's own PHP engine; a site with no plugins runs no PHP at all.

Where we are careful

Three things that travel with every number here.

The point of publishing the stacks is that a reader can check them. So here is what the numbers above do not say.

WordPress has a sub-millisecond best case

With a page cache in front of it, WordPress answers an anonymous repeat visit in a fraction of a millisecond: nginx returning a stored file, with PHP never running. That is real, and it is faster than any page that has to be rendered. It ends where its permission ends: no cart, no checkout, no search, nothing logged in. Every number here is WordPress executing, which is the case a cache cannot cover.

Throughput under full load is being re-measured

The numbers on this page are times for one page at a time. Our last full-load measurement was taken on an older release and no longer describes the current engine. We are re-measuring pages per second at saturation and will publish the result here.

Classic PHP themes are newer ground

WordPress's own classic templates are lighter than its block themes, and BookPress's classic-theme rendering is more recent than its block-theme path: a few of the bundled classic themes still fall back to an error today. The content numbers above are block themes, which is where WordPress core itself is heading.

Check the numbers yourself

The demo is a real WordPress site on the smallest machine Google Cloud sells, with no page cache in front of it. Sign in, publish a post, open the shop, ask for a page no cache could have kept.

Open the live demo