Zstd Compressor

Zstandard is a modern lossless compression algorithm developed at Meta. It reaches compression ratios close to gzip at its best settings while running several times faster, which is why the Linux kernel, package managers and big-data pipelines have adopted it as their default format (.zst).

No. Compression and decompression run entirely inside your browser tab on a WebAssembly build of the official zstd codec. Your files never leave your device, and the tool keeps working once the page has loaded even without a connection.

Levels 1–3 are extremely fast and already shrink most files substantially; level 3 is the zstd default and the best all-round choice. Levels 10–19 squeeze out extra bytes but take noticeably longer and use more memory — level 19 is meant for archiving where time does not matter.

Files up to 512 MB are accepted. Processing happens in a background worker thread so the page never freezes, but the memory available to your browser is the practical ceiling for very large inputs.

A .zst frame stores only the compressed bytes and no original file name. The tool therefore names the restored file after the uploaded archive with a .out extension — just rename it back if you know what the original was called.

Yes. The output is a standard zstd frame, byte-compatible with the official zstd command line at the same level. Use `zstd -d file.zst` on Linux, macOS or Windows, or any archive manager with zstd support.

Against gzip, zstd is usually several times faster at similar or better ratios. ZIP is an archive container that bundles many files with metadata, while .zst compresses a single data stream — use zip for folders and zstd when you want maximum speed and ratio for one file.

Data that is already compressed — JPEG photos, videos, ZIP archives — contains almost no redundancy, so zstd has nothing to remove and the small format overhead can even enlarge the file slightly. The tool always shows exact before/after sizes so you can keep whichever file is smaller.