Compress & Archive

Compression, both directions, on your own machine

This is the tool that does not care what your file is. It works on bytes: point it at a log, a CSV export, a database dump, an SVG or a folder of source files, and it will shrink them with the same algorithms your web server and your build pipeline already use. Drop a .gz or a .zip back in and it reads them just as happily.

Which codec to reach for

gzip

Understood by everything. The safe default for anything you will ship.

Brotli

Noticeably smaller than gzip on text. Every current browser can read it.

zlib

DEFLATE with a zlib header. What PNG chunks and many protocols embed.

DEFLATE

Raw stream, no header or checksum. Only useful if something expects it.

ZIP

A container, not a stream: bundles many files into one archive.

What the level slider buys you

Higher levels spend more time searching for repetition before giving up, so the file gets smaller and the wait gets longer. The curve flattens fast: gzip 9 is only a few percent smaller than gzip 6 on most inputs, while Brotli keeps finding wins up to 11 because its window reaches four megabytes back. Decompression speed is unaffected by the level — you only pay once, when you compress.

What will not shrink

JPEG, PNG, WebP, AVIF, MP3, MP4 and ZIP files have already had their repetition squeezed out. Running them through gzip usually costs a few bytes of header rather than saving anything, and the result panel will show that as a positive percentage. If you need those smaller, re-encode them in the Image Studio instead — that changes the pixels, which is where the real savings are.

Frequently asked questions

Which should I use, gzip or Brotli?

Brotli, if whatever reads the file supports it: on text it is usually 15-25% smaller than gzip at a comparable level. Choose gzip when the consumer is old, unknown, or a command-line tool you do not control.

Can I make a ZIP file in the browser?

Yes. Drop as many files as you like, switch the output to ZIP, and they are bundled into a single archive with DEFLATE compression. Entry names are stripped of directory-traversal segments before they are written.

Can it open an existing ZIP?

It can. Drop the archive in extract mode to read its table of contents — names, original size and compressed size — and pull out individual entries without unpacking the whole thing.

Why did compressing my JPEG barely help?

Because it is already compressed. gzip and Brotli look for repetition, and a JPEG, MP4 or ZIP has had most of its repetition removed already. They pay off on text, code, logs, CSVs, SVGs and uncompressed formats like BMP or WAV.

Is there a file-size limit?

Only what your browser will hand a single tab. Compression is done in memory in a Web Worker, so multi-hundred-megabyte files are practical on a desktop and considerably less so on a phone.