bzip2

Updated

September 4, 2026

Overview

bzip2 compresses files with the Burrows–Wheeler algorithm (.bz2). It typically achieves better ratios than gzip but is slower and has largely been superseded by xz (better ratio) and zstd (better speed/ratio balance) for new work. You will still encounter .bz2 and .tar.bz2 in older archives and some software distributions.

Syntax

bzip2 [options] [file...]
bunzip2 [options] [file...]
bzcat [file...]

Common Options

Option Description
-c Write to stdout
-d Decompress
-z Compress (default)
-k Keep input files
-f Force
-t Test integrity
-v Verbose
-q Quiet
-s Small memory mode
-1 … -9 Block size / effort (100k–900k blocks)

Examples with Explanations

Compress / decompress

bzip2 file.txt
bzip2 -k file.txt
bunzip2 file.txt.bz2
bzip2 -d file.txt.bz2

Pipelines

bzip2 -c file.txt > file.txt.bz2
bzcat file.txt.bz2 | less
tar cjf archive.tar.bz2 dir/
tar xjf archive.tar.bz2

Levels

bzip2 -1 quick.dat
bzip2 -9 small.dat

Test

bzip2 -t file.txt.bz2

Search compressed

bzgrep 'ERROR' app.log.bz2
bzless app.log.bz2

Prefer modern tools when choosing

# old style
tar cjf backup.tar.bz2 data/
# modern defaults for new archives
tar --zstd -cf backup.tar.zst data/
# or
tar cJf backup.tar.xz data/

Notes / Pitfalls

  • Deletes inputs by default without -k.
  • CPU-heavy relative to gzip/zstd for similar tasks.
  • Memory scales with block size (-s helps constrained systems).
  • Less common in new HTTP APIs than gzip/br/zstd.
  • Parallel variant: pbzip2 / lbzip2 if installed.

2026-relevant notes

  • Maintain decompression support for legacy artifacts; avoid bzip2 for new internal pipelines unless required.
  • GNU tar still supports -j / --bzip2.
  • If an upstream only publishes .tar.bz2, trust tar xjf / bzip2 -d rather than recompressing without need.

Additional Resources

  • man bzip2