How does the HTTP compression test work?

Our server asks the website for the URL you enter twice, using HEAD requests, which fetch only the response headers, not the file itself:

1. once without asking for compression, and
2. once saying it accepts the compression methods our server supports (gzip, deflate, Brotli and zstd).

If the second response has a Content-Encoding header, the file is compressed, and that header names the method. The sizes come from each response's Content-Length header, and the savings are the difference between them.

Many servers don't send a size for compressed responses, because they compress the file as they send it. The size then shows as Unknown, even when compression is on.

Only public http:// and https:// addresses can be checked, and redirects aren't followed, so the result is for the exact URL you enter. Each visitor can run up to 20 checks a minute.

The line at the bottom about your own browser comes from the Accept-Encoding header your browser sends with every request, listing the compression methods it understands.

Last updated: September 29, 2026 (12:47 AM GMT)