7-Zip is a file archiver with a high compression ratio. Usually, 7-Zip compresses to 7z format 30-70% better than to zip format. High compression ratio in new 7z format with LZMA compression. Supported formats - Packing / unpacking: 7z, ZIP, GZIP, BZIP2 and TAR. Unpacking only: 7z, ZIP, RAR, GZIP, GZ, BZIP2, BZ2, TAR, CAB, ISO, ARJ, LZH, LHA, CHM, MSI, WIM, Z, CPIO, RPM, DEB, CPIO, DMG, FAT, HFS, LZMA, NTFS, SPLIT, SWM, TAZ, TBZ, TBZ2, TGZ, TPZ, VHD, XAR, XZ, Z and NSIS. For ZIP and GZIP formats, 7-Zip provides a compression ratio that is 2-10 % better than the ratio provided by PKZip and WinZip. Self-extracting capability for 7z format. Integration with Windows Shell. Powerful File Manager. Localizations for 70 languages.

Powerful command line version
Plugin for FAR Manager
7-Zip works in Windows XP, Vista, 7, 8, 10 (32/64-bit). There is a port of the command line version to Linux/Unix.
Compression ratio results are very dependent upon the data used for the tests.

The main features of 7z format:
Open architecture
High compression ratio
Strong AES-256 encryption
Ability to use any compression, conversion or encryption method
Supports files with sizes up to 16000000000 GB
Unicode file names
Solid compression
Archive header compression

7z has an open architecture, so it can support any new compression method, too. The following methods currently are integrated into 7z:
LZMA - Improved and optimized version of LZ77 algorithm
PPMD - Dmitry Shkarin's PPMdH with small changes
BCJ - Converter for 32-bit x86 executables
BCJ2 - Converter for 32-bit x86 executables
BZip2 - Standard BWT algorithm
Deflate - Standard LZ77-based algorithm

LZMA is the default and general compression method of 7z format. The main features of LZMA are:
High compression ratio
Variable dictionary size (up to 4 GB)
Compression speed: about 1 MB/s on 2 GHz CPU
Decompression speed: about 10-20 MB/s on 2 GHz CPU
Small memory requirement for decompression (depend from dictionary size)
Small code size for decompression: about 5 KB
Supports multi-threading and P4's hyper-threading