GZIP Compress & Extract

How to compress or extract GZIP

1. 1

   Choose Compress or Extract

   Compress makes a .gz from one file. Extract turns a .gz back into the original file.

2. 2

   Choose a file

   Pick one file or drop it on the page. Compress also needs a preset; Extract suggests an output name you can edit.

3. 3

   Process and download

   Press the button, check the sizes, then download the result.

What to know

• One file in, one file out. GZIP compresses a single stream of bytes. It does not store file names, folders or several files like a ZIP does. To bundle files, use Create ZIP.
• Fast, Balanced or Maximum. These use compression levels 1, 6 and 9: higher levels are slower and usually a little smaller. The result shows the real original size, .gz size and change, and says when the file did not shrink.
• Checked on the way out. When you extract, the data is compared with the CRC-32 checksum and length stored in the GZIP file. If they do not match, or the file is cut off or has extra bytes after its end, you get an error and no file.
• Joined GZIP files. A .gz made of several GZIP members joined together (what cat a.gz b.gz makes) is extracted as one file, with every member checked.
• .tar.gz and .tgz. These are a TAR archive compressed with GZIP. The tool removes the GZIP layer and gives you a .tar file. It does not extract TAR, so the files inside are not opened.
• Names are suggestions. The output name comes from the file you chose (data.csv.gz becomes data.csv) and you can edit it. A name stored inside the GZIP is untrusted: it is cleaned, used only if there is nothing better, and never treated as a path.

Limits

largest file to open

1 GB

largest extracted result

512 MB

A result that would expand more than 250× once it passes 64 MB is stopped as a possible decompression bomb, as is a run longer than 10 minutes. The finished file is held in your browser until you download it. Work runs in a background thread (a Web Worker) and Cancel stops it at once.

Frequently asked questions