zpaq

Create archive

$ zpaq add [archive.zpaq] [files...]

$ zpaq extract [archive.zpaq]

$ zpaq list [archive.zpaq]

$ zpaq add [archive.zpaq] [files] -method 5

$ zpaq extract [archive.zpaq] -to [/output/dir]

$ zpaq add [backup.zpaq] [directory] -all

zpaq command archive[.zpaq] [files...] [options]

zpaq is an archiver with high compression ratios and journaling capability. It uses context mixing and neural network-based prediction for compression, achieving ratios better than most conventional archivers.
ZPAQ's journaling feature makes it excellent for incremental backups. Each add creates a new version, and previous versions remain accessible. The -until flag extracts files as they existed at a specific time.
Archives are append-only by default - new versions add to the archive without modifying previous data. This provides some protection against corruption and allows versioned backups.
Compression levels range from 0 (store only) to 5 (maximum compression, slow).

-method level

Compression level (0-5, higher = better compression)

Destination directory for extraction

Include hidden files and empty directories

Exclude files matching pattern

Include only matching files

Extract version as of date

Number of threads to use

Encrypt/decrypt with password

add

Add or update files in archive

Extract files from archive

List archive contents

Compare archive with filesystem

Test archive integrity

High compression levels are CPU-intensive and slow. Level 5 can be orders of magnitude slower than level 1.
ZPAQ format is not widely supported outside the zpaq tool itself. Consider compatibility needs.
Journaling archives can grow large over time. Use zpaq add -method 0 to consolidate.
Encryption is optional and must be specified during archive creation.

tar(1), 7z(1), gzip(1), zstd(1)