Gzip compress

It depends entirely on the content. Text, CSV, JSON and logs usually shrink a lot; already compressed data such as JPEG, MP4, or a zip barely moves and can come out slightly larger, since gzip adds a header and a trailer. The tool compresses with fflate gzipSync and reports the result without making any promise about the ratio.

No. There are no options at all, so every run uses the fflate default. That keeps the result the same every time you compress the same bytes. If you need maximum compression, use a desktop tool that exposes the level.

Your file name with .gz appended, keeping the original extension, so report.tar becomes report.tar.gz and notes.txt becomes notes.txt.gz. Run Gzip decompress on it later and the .gz is stripped back off. Nothing about the original name is thrown away, so you can always see what the compressed file used to be.

Not in one file. gzip compresses a single stream, so there is no directory structure inside a .gz. Bundle the files first with Create tar, then gzip the tar, which is exactly how a .tar.gz is made. That is why a .tar.gz carries two extensions: one for the bundle and one for the compression.

Yes. The tool works per file, so drop as many as you like and each comes back as its own .gz, named from its own source. They are compressed one after another and listed separately for download. Nothing is bundled together for you, so use Create tar or Create ZIP first if you want a single file at the end.

No. Compression happens in a Web Worker in your own tab and nothing is sent to a server. The input and the compressed output are both held in memory at once, so a large file is bounded by what your browser tab can hold rather than by any rule in the tool.