What is JXL (JPEG XL)?

How JPEG XL Works

JPEG XL is unusual because it contains two distinct compression engines under one roof. For photographic content it uses a variable block discrete cosine transform, often called VarDCT, which extends the familiar DCT idea of the original JPEG but allows the block size to vary across the image. Smooth areas can use larger transforms while detailed areas use smaller ones, which spends bits where the eye actually notices them. For graphics, screenshots, and lossless work it uses a separate modular mode that grew out of earlier research codecs in the same lineage.

One of the most practical features is lossless JPEG recompression. A great deal of the world's stored imagery already exists as ordinary JPEG files, and JPEG XL can take such a file and repack it, typically around twenty percent smaller, in a way that can be reversed to reproduce the exact original bytes. This lets an archive shrink without discarding anything or committing to a one way conversion, which is a strong migration argument.

The format supports progressive decoding, so a partially downloaded file can already show a coherent lower quality image that sharpens as more data arrives. It also handles wide color gamut, high dynamic range, high bit depths, alpha transparency, and animation, meaning a single format can cover roles that today are split across JPEG, PNG, and GIF.

Because the same file can carry lossy or lossless data and can encode either photographs or flat graphics well, JPEG XL was designed as a general purpose still image format rather than a specialist tool. Its container also supports embedded color profiles and metadata so that color managed workflows survive round trips.

History and Standardization

JPEG XL was developed by the Joint Photographic Experts Group committee as a modern successor to the original JPEG, drawing on two earlier projects whose ideas were merged into a single design. It is published as an ISO standard, ISO/IEC 18181, with the core parts reaching final status around 2021 and 2022. The .jxl extension identifies these files.

A deliberate goal of the project was to be royalty free, removing the licensing uncertainty that had slowed some earlier image standards. The committee positioned it explicitly as a format that could eventually take over the roles of JPEG, PNG, and GIF at once, backed by the lossless JPEG transcoding path to ease the transition from existing libraries.

Adoption has been shaped mostly by browser politics rather than technical merit. Support arrived behind experimental flags in some browsers and was later removed from at least one major browser, while others have varied in their commitment. This uneven native browser support is the main reason JPEG XL has not yet become a default web format despite broad praise for its capabilities.

JPEG XL Compared to Other Formats

Against the original JPEG, JPEG XL offers better quality per byte, a real lossless mode, transparency, and the transcoding trick that lets old JPEGs shrink reversibly. The catch is reach: JPEG is understood by essentially every device and browser ever made, while JPEG XL decoding is not guaranteed everywhere, so JPEG remains the safe default for maximum compatibility.

Compared to AVIF, which is built on the AV1 video codec and is also royalty free, the two are close rivals. AVIF tends to be very strong at low bitrates and benefits from wide browser support today, while JPEG XL is often praised for high quality and lossless work and for its progressive decoding and JPEG recompression. The gap is small enough that browser availability usually decides which one a project actually uses.

Against WebP and PNG, JPEG XL is technically more capable and more flexible, covering both lossy and lossless roles with better efficiency. The honest limitation is deployment: until native support is dependable across the browsers and tools your audience uses, PNG and WebP remain the pragmatic choices for the web even where JPEG XL would produce smaller files.