JXL Format Guide

JPEG XL (JXL) is a next-generation image format standardized as ISO/IEC 18181, developed by the JPEG committee to serve as a universal replacement for legacy JPEG and other image formats. The format combines innovations from Google's PIK proposal and Cloudinary's FUIF (Free Universal Image Format) into a single, comprehensive codec that supports both lossy and lossless compression. JPEG XL achieves approximately 60% better compression than JPEG at equivalent visual quality, and outperforms PNG, WebP, and AVIF across a wide range of image types. The format supports images with up to 32 bits per channel, enabling true HDR (High Dynamic Range) imaging with wide color gamut coverage including Rec. 2100 and Display P3 color spaces. One of JPEG XL's most distinctive features is its ability to losslessly recompress existing JPEG files, reducing their size by approximately 20% while allowing bit-exact reconstruction of the original JPEG, making it ideal for migrating existing photo libraries without any quality loss.

History of JPEG XL

JPEG XL development began in 2017 when the JPEG committee issued a call for proposals for a next-generation image coding standard. The project received seven responses, with two proposals standing out: PIK from Google (led by Jon Sneyers and Jyrki Alakuijala) and FUIF from Cloudinary (led by Jon Sneyers). In 2018, the committee decided to merge these two proposals into a single format, combining PIK's advanced entropy coding and perceptual optimization with FUIF's responsive-by-design progressive decoding architecture. The reference implementation, libjxl, was developed as an open-source project and released under the BSD 3-Clause license. The format specification was published as ISO/IEC 18181 in March 2022, with the standard consisting of four parts covering the core codestream, file format, conformance testing, and reference software. Browser support has been contentious: Chrome added experimental JXL support behind a flag in version 91 (2021) but later removed it in version 110 (2023), sparking significant community backlash. Safari added JPEG XL support in iOS 17 and macOS Sonoma (2023), making Apple the first major platform vendor to ship native JXL support. As of 2025, JPEG XL is supported by many image editors, viewers, and processing tools including Photoshop, GIMP, ImageMagick, darktable, and the Pillow Python library.

Key Features and Uses

JPEG XL's feature set is remarkably comprehensive. The format supports progressive decoding, allowing images to render at increasing quality levels as data arrives, which is crucial for web delivery over slow connections. Its modular design enables both simple and complex encoding strategies: a lightweight "modular" mode provides fast encoding suitable for lossless compression, while the full VarDCT (Variable-size Discrete Cosine Transform) mode delivers state-of-the-art lossy compression. JPEG XL natively supports animation with frame-level encoding, offering a modern replacement for animated GIF and WebP with vastly superior compression. The format includes built-in support for alpha transparency, depth maps, multiple frames, and extra channels for arbitrary data such as thermal or spectral imaging. JPEG XL preserves rich metadata including EXIF, XMP, and JUMBF, and supports embedded ICC color profiles for accurate color management. The royalty-free licensing ensures JPEG XL can be freely implemented by anyone, addressing a key concern that has historically fragmented the image format ecosystem.

Common Applications

JPEG XL is increasingly used for photo storage and archival, where its lossless JPEG recompression feature allows photographers and libraries to reduce storage costs by 20% without altering their original JPEG files. Web developers adopt JPEG XL for image delivery where supported, as its superior compression ratios translate directly to faster page loads and reduced bandwidth costs. The format's HDR support makes it valuable for professional photography, digital cinema, and display manufacturing workflows that require accurate representation of high dynamic range content. Print and publishing industries benefit from JPEG XL's ability to handle large, high-resolution images with efficient compression while preserving color accuracy through ICC profiles. Scientific imaging applications leverage the format's support for high bit depths and extra channels to store multispectral, thermal, and medical imaging data. Game development studios use JPEG XL for texture storage and asset distribution, as the format's compression efficiency reduces download sizes for game updates and patches. Converting from JPEG XL to widely supported formats like JPG, PNG, or WebP remains essential for compatibility with platforms and browsers that do not yet support JXL natively.

Advantages and Disadvantages

Advantages

Superior Compression: ~60% smaller than JPEG at equivalent quality, outperforms PNG and WebP
Lossless JPEG Recompression: Reduce JPEG file sizes by ~20% with bit-exact reconstruction
HDR and Wide Gamut: Up to 32 bits per channel with Rec. 2100 and Display P3 support
Progressive Decoding: Images render at increasing quality levels during download
Both Lossy and Lossless: Single format handles all compression needs
Animation Support: Modern animated image format replacing GIF and animated WebP
Royalty-Free: Open standard with BSD-licensed reference implementation
Rich Metadata: Full EXIF, XMP, ICC profile, and JUMBF metadata support

Disadvantages

Limited Browser Support: Not supported in Chrome or Firefox; Safari-only among major browsers
Slow Encoding: High-quality encoding is computationally expensive and slower than JPEG
Complex Implementation: Codec is far more complex than simpler formats like QOI or PNG
Ecosystem Fragmentation: Competing with AVIF and WebP creates format confusion
Large Decoder Size: libjxl binary size is significant for embedded or web-based decoders
Inconsistent Platform Support: Chrome removed support, creating adoption uncertainty
Encoding Memory Usage: High-quality encoding requires substantial RAM for large images
Still Maturing: Tooling and ecosystem less established than JPEG, PNG, or WebP