File extension

.lz4

LZ4 frame format

.lz4 files are LZ4-compressed streams: a lossless format built for very high compression and decompression speed, trading some compression ratio.

Known meanings

Formats and reference material for .lz4

LZ4 frame format

Container for LZ4 block-compressed data with optional checksums and size fields

  • archive
  • Yann Collet
Open link

LZ4 block format

Raw LZ4 compressed block layout used inside frames and by embedded users of the library

  • archive
  • Yann Collet
Open link

Details

Format notes

LZ4

LZ4 is a lossless compression algorithm by Yann Collet that prioritizes speed. It compresses and decompresses at very high rates, often limited by memory or storage bandwidth rather than by the CPU, at a lower compression ratio than algorithms such as gzip or xz. The same author later created Zstandard.

Technical Notes

LZ4 is a byte-oriented LZ77-style scheme: data is encoded as sequences, each with a token, optional literal bytes, and a match described by an offset and length, with no separate entropy coding stage. This keeps decoding simple and fast. The .lz4 file extension normally refers to the LZ4 frame format, which wraps a series of blocks with a header and optional content size, block and content checksums, and an end mark. A frame begins with the magic number 0x184D2204, stored little-endian as the bytes 04 22 4D 18. The frame format and the underlying block format are both documented in the project repository. A high-compression variant, LZ4HC, produces smaller output while keeping the same decompression speed.

Adoption

LZ4 is used in file systems, databases, in-memory caches, game engines, the Linux kernel, and network and storage software where low latency matters. The reference implementation provides a library and a command-line tool, and bindings exist for many languages.

Preservation And Security Notes

LZ4 files are well suited for fast, transient compression; for long-term archival, a higher-ratio format such as xz or Zstandard may be preferable. Use the frame checksums where possible, since the block format itself has no integrity check. Decoders handling untrusted data should bound output size and use a maintained library version.

Further Reading

  • LZ4 project: https://github.com/lz4/lz4
  • Frame format: https://github.com/lz4/lz4/blob/dev/doc/lz4_Frame_format.md
  • Block format: https://github.com/lz4/lz4/blob/dev/doc/lz4_Block_format.md