Online .TAR.ZST Archive Unpacker
(No Size Limits)

A .tar.zst file is two formats stacked. A TAR archive bundles many files into one without compressing anything, then Zstandard compresses that bundle. You will also see the same thing named .tzst. Both layers have to come off, in that order, which is why tools that only understand one of them report a corrupt or unrecognised file.

Why your computer probably cannot open it

Zstandard is young. Yann Collet released it at Facebook in 2016 and it was standardised as RFC 8878 in 2021, decades after ZIP and gzip became universal. Linux adopted it fast: Arch Linux moved its entire package repository to .pkg.tar.zst around 2020, and Fedora, Debian and btrfs followed. Desktop tooling did not keep pace.

System	Opens it natively	What to do
Windows 11 (23H2 and later)	Yes	Right-click, Extract All
Windows 10	No	Install 7-Zip, or extract on this page
macOS (any version)	No	brew install zstd, or extract on this page
ChromeOS	No	Extract on this page, since nothing is installable
Linux	Usually	tar --zstd -xf

The result is a format that is routine in Linux and developer work, and awkward almost everywhere else.

Why Zstandard rather than gzip

The gain is not compression ratio, which is broadly comparable to gzip at default settings. It is speed. Zstandard decompresses several times faster than gzip at similar ratios, and it scales across a wide range of levels, so a distribution can compress packages hard once and have every user decompress them cheaply. For anything downloaded far more often than it is built, that trade is worth making, which is exactly why package repositories moved.

Format	Compression ratio	Decompression speed	Where you meet it
.tar.gz	Moderate	Moderate	Source tarballs, the long-standing default
.tar.xz	Highest	Slowest	Kernel sources, where download size matters most
.tar.zst	Comparable to gzip, higher at upper levels	Fastest of the three	Arch and Fedora packages, container layers, backups

How this page extracts it

The archive is read with file.stream() and piped through a Zstandard decompression stream, and the TAR entries are written to disk as they emerge. Nothing is ever fully buffered, so memory stays flat whether the archive is 5 MB or 5 GB, and none of it leaves your machine. There is no upload step to wait through and no server copy to trust.

Detection does not rely on the filename. A Zstandard frame starts with the magic bytes 28 B5 2F FD, which are checked directly, so a mislabelled or extension-less file still resolves correctly.

The 10 MB ceiling that breaks large .zst files elsewhere

This is worth documenting, because it silently affects other browser-based extractors. A widely used WebAssembly Zstandard decoder contains this check:

if (e > 10000000) throw new err('win 2 large');

The wording suggests a window-size limit, but the value being tested is Frame_Content_Size, the declared size of the decompressed output. The effect is a hard 10 MB cap on Zstandard output regardless of how the file was compressed. A 33 MB SQL dump fails on it. So does any real archive.

unpk.app decodes .tar.zst with libzstd, the reference implementation, driven through libarchive. There are two caps to get past here, and neither is in the file: that 10 MB output check, and the 128 MB window limit almost every build of libzstd ships with by default, which is why zstd -d itself tells you to pass --memory=2048MB on some archives. Ours is set to the largest window the format defines, so a .tar.zst compressed with --long or --ultra decodes here without being asked to justify itself. If a .tar.zst has failed for you elsewhere with a vague error, one of those two caps is the likely cause.

Doing it from a terminal instead

If you have a shell and do not mind installing things, GNU tar handles Zstandard directly:

tar --zstd -xf archive.tar.zst

Older versions of tar predate the --zstd flag, so pipe it through the zstd binary instead:

zstd -d archive.tar.zst -c | tar -x

On macOS neither works until you install the tool, with brew install zstd. That install step is the reason this page exists: on a managed laptop, a school Chromebook or someone else's machine, you often cannot run it.

What does not work

Multi-volume archives split across several files (.zst.001 and similar) are not supported, and neither is encrypted 7z. Zstandard itself has no encryption layer, so a password-protected .tar.zst is really an encrypted container holding one, and that outer layer has to come off first.