ZIP Creator

A ZIP archive compresses each entry independently and records them all in a directory written at the end of the file. That structure is why a single file can be extracted without decompressing everything before it, and why a ZIP can be built incrementally. The trade-off is that per-file compression cannot exploit similarity between files: a hundred near-identical documents compress no better together in a ZIP than separately, whereas a solid archive format such as tar.xz compresses the concatenated stream and does dramatically better on exactly that case.

Compression only helps where there is redundancy left to remove. JPEG, PNG, MP4, MP3 and PDF are already compressed, and zipping them yields a file within a percent or two of the original size, occasionally slightly larger, once the archive overhead is counted. Text, CSV, JSON, XML, source code and uncompressed bitmaps compress enormously, often to a fifth of their size or less. If the goal is a smaller download, the composition of the input decides whether ZIP can help at all; if the goal is bundling many files into one, it helps regardless.

ZIP provides no meaningful confidentiality and should not be used for it. Entry names, sizes and timestamps sit in the directory in plain text even when the contents are encrypted, so the archive reveals its own structure to anyone holding it. The original ZipCrypto encryption is broken well enough to be treated as absent. If contents need protecting, encrypt them before archiving with a tool built for it. Archives are created here in your browser, so the files being bundled are never uploaded, which is the property worth having when they are documents you were sent rather than ones you published.