JPEG XL: The Next Evolution in Image Formats

Look, we've been through this dance before. PNG was supposed to replace GIF. WebP was supposed to replace JPEG. AVIF was supposed to replace everything. And yet here we are in 2026, still mostly using JPEG for photos and PNG for graphics, just like we did in 2006.

But JPEG XL might actually be different. Not because it's technically superior (though it is), but because it was designed to solve real problems that actual humans have with image formats. And that's a rare thing in the world of technical standards.

What makes JPEG XL different?

First off, the compression is legitimately impressive. In independent tests, JPEG XL consistently produces files that are 20-40% smaller than equivalent-quality JPEG images. That's not a marginal improvement — that's the kind of difference that actually matters when you're trying to load a photo-heavy webpage on a phone with spotty reception.

But here's what's really clever: JPEG XL can take your existing JPEG files and recompress them losslessly into smaller files. You read that right. You can convert a 5MB JPEG into a 3.5MB JPEG XL, then convert it back to JPEG later and get the exact same file. Pixel-for-pixel identical.

This is huge for migration. With previous formats, converting your library meant choosing between keeping the originals around forever (wasting space) or accepting quality loss from re-encoding. JPEG XL eliminates that tradeoff.

The feature set actually makes sense

JPEG XL supports a bunch of features that sound boring in a spec document but are genuinely useful in practice:

• Progressive decoding: The image renders quickly at low quality, then improves as more data loads. Like progressive JPEG but way better.
• High bit depth: 16-bit and 32-bit per channel support for HDR and professional photography work.
• Animation: Yes, it can replace GIF and compete with WebP animations. With better compression and quality.
• Lossless mode: Better compression than PNG for photos (about 35% smaller), while still being completely reversible.
• Fast encoding/decoding: The format was designed to be computationally efficient. It's not slow like AVIF.

The format can also handle transparency, multiple layers, and even has built-in support for image regions with different quality settings (so you can compress the background more aggressively than the subject).

So what's the catch?

Browser support. The eternal problem with new image formats.

As of April 2026, Safari supports JPEG XL natively (Apple added it in Safari 17). Firefox has support, but it's behind a flag you need to enable manually. Chrome? That's complicated. Google added support, then removed it in 2022, then started working on re-adding it in 2024 after significant backlash. Experimental support is back in Chrome Canary, but it's not in stable releases yet.

This fragmented support means you can't just drop JPEG XL files into your website and call it a day. You need fallbacks. The standard approach is using the <picture> element:

<picture>
  <source srcset="photo.jxl" type="image/jxl">
  <source srcset="photo.webp" type="image/webp">
  <img src="photo.jpg" alt="Description">
</picture>

Browsers will use the first format they understand, falling back to good old JPEG if necessary. It works, but it's extra work.

Real-world testing: the numbers

I converted about 200 photos from my library to compare formats. Here's what I found (average file sizes for equivalent perceptual quality):

• Original JPEG (quality 85): 2.8 MB
• WebP (equivalent quality): 2.1 MB (25% smaller)
• AVIF (equivalent quality): 1.9 MB (32% smaller)
• JPEG XL (equivalent quality): 1.8 MB (36% smaller)

JPEG XL came out ahead, but only by a small margin over AVIF. The real difference showed up in encoding time. JPEG XL was about 4x faster to encode than AVIF, which matters a lot if you're batch-processing hundreds of images or generating thumbnails on the fly.

For lossless compression of the same photos (maintaining perfect quality), JPEG XL averaged 4.2 MB per image, while PNG was 6.5 MB. That's 35% savings for exactly the same pixels.

Who should care about this format?

If you're a photographer managing a large library, JPEG XL is worth experimenting with now. The lossless recompression feature alone could save you substantial storage space without any quality loss. And when browser support improves, you'll already have an optimized library ready to go.

For web developers, it's more of a "wait and see" situation. You can implement JPEG XL with fallbacks today, but the complexity may not be worth the 10-15% extra file size reduction over WebP. Check your analytics — if Safari users are a significant portion of your audience, it might be worth it.

E-commerce sites with lots of product photos could benefit significantly. Faster page loads mean better conversion rates, and even small file size improvements add up when you're serving millions of images per day. Tools like KokoConvert's image compression can help test JPEG XL against your current formats.

Converting to JPEG XL

If you want to experiment, conversion is straightforward. The reference implementation (libjxl) has command-line tools, and many image processors now support the format. For web workflows, you can convert images to JPEG XL directly in your browser without uploading anything.

For batch conversion of large libraries, here's what works well:

• Keep originals in a separate folder (trust me on this)
• Use quality settings between 85-95 for photos (90 is a good default)
• Test a few images before converting everything
• Use lossless mode for archival purposes or when you might need to convert back

And remember: you can always convert JPEG XL back to JPEG later if browser support doesn't pan out the way we hope. That's the whole point of the lossless recompression feature.

The bigger picture

JPEG has been the dominant photo format since 1992. That's 34 years. It's outlasted Netscape, Internet Explorer, Flash, and countless other web technologies that once seemed permanent. For a new format to actually replace it will take time — probably a decade or more.

But JPEG XL has something most previous challengers didn't: a practical migration path. The fact that you can losslessly recompress existing JPEGs means adoption doesn't require destroying and recreating existing image libraries. That's a big deal.

Is it ready for production use everywhere? Not yet. But it's worth keeping an eye on, especially if you work with images professionally. The technical foundation is solid, and the features actually solve real problems.

Sometimes that's enough.