zstd

Compress a file

$ zstd [file]

$ zstd -d [file.zst]

$ zstd -k [file]

$ zstd -19 [file]

$ zstd --ultra -22 [file]

$ zstd -1 [file]

$ zstd -c [file] > [file.zst]

$ zstd -dc [file.zst]

$ zstd -T0 [file]

$ zstd -t [file.zst]

zstd [options] [files...]

Zstandard (zstd) is a fast lossless compression algorithm providing high compression ratios. It offers a wide range of compression/speed trade-offs, from ultra-fast to ultra-compression modes.
At similar compression ratios, zstd is typically faster than zlib/gzip and significantly faster than xz/lzma. Higher compression levels approach xz ratios while maintaining faster decompression.
zstd supports dictionary compression for small files, parallel compression, and can serve as a drop-in replacement for gzip in many workflows. The format is defined in RFC 8478.

-d, --decompress

Decompress files.

Keep source files.

Write to stdout.

Compression level (default 3).

Enable ultra compression (levels 20-22).

Number of threads (0=auto).

Test compressed file integrity.

List compressed file information.

Force overwrite/compression.

Recurse into directories.

Remove source file after compression.

Use dictionary for compression.

Create dictionary from samples.

Suppress output.

Verbose mode.

Higher compression levels use more memory. Ultra levels (20-22) require significantly more memory. Dictionary compression requires same dictionary for decompression. Not all tools support .zst format yet.

Zstandard was developed by Yann Collet at Facebook starting in 2015, building on his earlier LZ4 work. It was released publicly in 2016 and became RFC 8478 in 2018. Zstd has been adopted by the Linux kernel for compression, and is supported by tar, HTTP content-encoding, and numerous other tools.

gzip(1), xz(1), lz4(1), bzip2(1)