Gzip/Brotli Compression Test

Compare response size with Gzip/Brotli and without compression.

A brief history

gzip is a GNU utility (1992) based on the DEFLATE algorithm (RFC 1951/1952). It became the de-facto standard for compressing text web content in the HTTP/1.1 era.

Brotli is a modern compression codec by Google (publicly released 2015). With a richer static dictionary and smarter entropy coding it often outperforms gzip for HTML/CSS/JS/SVG, especially at higher compression levels.

Compression is configured at platform level on supported shared-hosting plans. On VPS and dedicated servers the result depends on the web server, content types and selected configuration.

How the test works

Request #1: Accept-Encoding: gzip, br — we detect the encoding chosen by the server and measure real bytes.
Request #2: Accept-Encoding: identity — forced no compression, measure bytes.
We calculate savings and output headers (Content-Encoding, Content-Type, Server).

Frequently asked questions

What exactly does the test do?

We send two requests: one with Accept-Encoding: gzip, br and one with Accept-Encoding: identity. We compare the actual downloaded bytes and show Content-Encoding plus savings.

Why can byte counts differ from Content-Length?

Content-Length may be missing (chunked) or reflect the already compressed size. We count the real downloaded bytes for each request.

Is Brotli always better than Gzip?

For text formats (HTML/CSS/JS/SVG) Brotli usually compresses better. For binary formats (JPEG/PNG/MP4/PDF) there is no benefit and they should not be recompressed.

Is compression enabled on HSTQ?

Compression is configured at platform level on supported shared-hosting plans. On VPS and dedicated servers it depends on the installed web server and selected management scope.

Gzip/Brotli