2026-03-24
AVIF vs WebP vs JPEG: Which Image Format Should You Use?
AVIF, WebP, and JPEG compared with an illustrative file-size example, artifact behavior, compatibility, and encoding trade-offs, plus a practical decision rule.

Last updated: June 27, 2026
Pick the smallest format your audience and delivery channels can display. For most modern sites that means testing AVIF first, retaining WebP as a broadly compatible fallback, and keeping JPEG where older clients or external systems require it. The file-size table below is one illustrative encode of the same 1600×900 photograph, not a universal savings guarantee.
Quick answer: AVIF, WebP, or JPEG?
Test AVIF as the primary format, serve WebP as a fallback, and keep JPEG for places modern formats cannot go (some email clients, legacy software, and partner feeds). AVIF often compresses photographs more efficiently, while the actual saving depends on the source image, encoder, settings, and quality metric.
A practical decision rule:
| Situation | Use | Why |
|---|---|---|
| Modern web pages, product photos, hero images | AVIF (+ WebP fallback) | Smallest bytes, best Core Web Vitals |
| Wide compatibility with one file, simple CMS | WebP | Native support in current major browsers and efficient compression |
| Email, RSS, old intranets, broken CMS pipelines | JPEG | Decodes everywhere, no negotiation |
| Sharp UI screenshots, text, line art | PNG or WebP (lossless) | JPEG and AVIF blur thin edges at low quality |
| Animations on a page | AVIF or animated WebP | Replaces GIF at a fraction of the size |
If your build pipeline cannot emit AVIF yet, switch to WebP first. It is the fastest single win, and you can layer AVIF on top later without changing the <img> markup.
What can the size difference look like?
The following 1600×900 feather-detail photograph illustrates one set of encoder outputs. Treat the numbers as an example to reproduce on your own images, not as a promised site-wide percentage:

| Format | Quality setting | Measured file size | Vs JPEG |
|---|---|---|---|
| JPEG | 85 | 222 KB | baseline |
| WebP | 85 | 171 KB | −23% |
| AVIF | 70 | 130 KB | −41% |
The result can change with image content and encoder settings. Noisy photographs, flat graphics, screenshots, and white-background product shots have different compression curves, so compare formats at an acceptable visual quality on a representative sample.

For a store with hundreds of product photos, even a modest per-image reduction can materially reduce transferred bytes. Measure the aggregate saving before adding another format to the build pipeline.
Does the smaller AVIF file actually look as good?
This is the question that matters, and the reason "AVIF is 40% smaller" is not the whole story. At sensible quality settings the answer is yes for photographs — and you can verify it on the same test image by zooming in:

At matched quality the differences are hard to see at normal viewing distance. The artifacts that do appear, when you push quality too low, are format-specific:
| Format | Failure mode when over-compressed | Where it shows first |
|---|---|---|
| JPEG | 8×8 blocking, ringing around edges | Skin tones, text, feather/fine detail |
| WebP (lossy) | Similar to JPEG but slightly cleaner at equal size | Same high-frequency areas |
| AVIF | Smooth smearing of fine texture, "plastic" look | Fur, foliage, film grain |
The practical lesson: keep AVIF in the 60–70 quality range for photographs. Below ~30, AVIF smears detail in a way that looks worse than a larger JPEG, because the human eye reads texture loss as "wrong" faster than it reads JPEG ringing.
How does each format compare on the dimensions that matter?
| Dimension | JPEG | WebP | AVIF |
|---|---|---|---|
| Year introduced | 1992 | 2010 | 2019 |
| Max compression | Lowest | Medium | Highest |
| Lossless support | No | Yes | Yes |
| Alpha (transparency) | No | Yes | Yes |
| HDR / wide gamut | No | Limited | Yes |
| Encoding speed | Fast | Fast (2–5× AVIF) | Slow |
| Compatibility | Broadest, including legacy systems | Current major browsers | Current major browsers; check external pipelines |
| Best for | Compatibility, email | General web | Photos, hero images |
Two trade-offs drive every format choice: AVIF costs more CPU to encode, and JPEG is the only format guaranteed to render in every mail client and legacy system. If you batch-encode once at upload time, the encoding cost is irrelevant. If you generate formats on the fly on every request, WebP's faster encoding matters.
When should you still pick JPEG?
JPEG is not obsolete — it is the universal fallback. Keep it for:
- HTML email. Most email clients strip or ignore WebP and AVIF. JPEG and PNG still rule here.
- Partner and marketplace feeds. Some product feeds, ad networks, and print pipelines only accept JPEG.
- Old embedded browsers (older Smart TVs, kiosks, e-readers) that predate WebP support.
- Small, already-optimized thumbnails where re-encoding to AVIF saves single-digit kilobytes and is not worth the pipeline complexity.
How do you serve all three without breaking old browsers?
Use the <picture> element with typed sources. The browser picks the first type it supports and ignores the rest.
<picture>
<source srcset="/img/product.avif" type="image/avif">
<source srcset="/img/product.webp" type="image/webp">
<img src="/img/product.jpg" alt="Green trail running shoe on white" width="800" height="600" loading="lazy">
</picture>
Notes that matter in practice:
- Always keep a real
<img>with a JPEGsrcas the final fallback. - Set
widthandheighton the<img>to prevent layout shift — this applies to every format. - Lazy-load below-the-fold images; do not lazy-load the LCP hero.
- If you use
next/image, it negotiates formats automatically when the loader and remote patterns are configured. This site'snext.config.tsalready allow-lists the CDN host.
How does format choice affect Core Web Vitals?
Images most often control Largest Contentful Paint (LCP) on image-heavy pages. Smaller bytes mean the hero image arrives and paints sooner. Field results track the file-size ratio closely:
| Format | Relative LCP | Notes |
|---|---|---|
| JPEG | Baseline | Largest bytes, slowest paint |
| WebP | ~20% faster | Good middle ground |
| AVIF | ~35% faster | Best when the hero is a photo |
Cumulative Layout Shift (CLS) is format-independent. It depends on whether you reserve space with width/height, not on whether the bytes are JPEG or AVIF. A badly-sized AVIF shifts the page just as much as a badly-sized JPEG.
Read Google's official guidance on images and Core Web Vitals and the image format reference for the current per-browser decoder support.
AVIF drawbacks worth knowing before you switch
AVIF is the right default for new work, but it has real rough edges:
- Encoding is slow. Single-threaded AVIF encoding can be 2–5× slower than WebP. Encode once at upload, not on every request.
- Older Safari (<16.0) and some embedded WebViews do not decode it. The
<picture>fallback covers this. - A few editors and DAM tools still cannot open AVIF. Keep a JPEG or PNG master.
- Very aggressive AVIF settings (quality < 30) smear detail. Stay in the 60–70 range for photos, as the zoom comparison above shows.
If any of these block your team, start with WebP. It removes the rough edges and still beats JPEG by ~23%.
A simple rollout order
- Measure first. Run PageSpeed Insights on your top templates and note image bytes and LCP.
- Add WebP as a fallback behind JPEG — quick win, no compatibility risk.
- Add AVIF sources above WebP in
<picture>for photographs. - Compress and resize every image to its display size before encoding. Use the Imagic AI Image Compressor and Image Converter for one-off or batch jobs.
- Reserve dimensions (
width/height) on every image to lock CLS. - Re-measure. Confirm LCP dropped and no image requests 404.
Common mistakes
- Serving one giant AVIF and skipping resize. Format does not save you from a 4000px image shown at 400px. Resize first, then encode.
- Lazy-loading the hero. The LCP image should load eagerly with
fetchpriority="high", not lazily. - Forgetting the
<img>fallback. A<picture>with only<source>tags and no<img>renders nothing on unsupported clients. - Comparing formats at the same quality number. AVIF q70, WebP q85, and JPEG q90 look roughly similar. Compare at matched visual quality, not matched numbers — that is exactly how the measurements above were taken.
- Pushing AVIF quality too low. Below ~30, AVIF's texture smearing looks worse than a larger JPEG.
Related guides
- Compress Images Without Losing Quality
- How Image Compression Works
- Image SEO Optimization Checklist
- Complete Image Optimization Checklist
- TinyPNG Alternatives Compared
Frequently asked questions
Is AVIF always smaller than WebP?
Not always; AVIF usually compresses photographs harder than WebP at matched visual quality, but flat graphics and screenshots can favor WebP or PNG depending on content.
How much smaller was AVIF than JPEG in this test?
In the measured encode above, AVIF q70 was 130 KB against a JPEG q85 baseline of 222 KB, a 41% reduction on that single image.
Do all browsers support AVIF?
Current major browsers decode AVIF, but older Safari below version 16 and some embedded WebViews do not, so a <picture> fallback to WebP or JPEG is still required.
Is WebP safe to use as the only format?
Yes; WebP has native support in current major browsers and beats JPEG by roughly 23% in the measured comparison above, making it a solid single-format choice if you cannot add AVIF yet.
When should I still use JPEG instead of AVIF or WebP?
Keep JPEG for HTML email, partner and marketplace feeds, and older embedded browsers that predate WebP and AVIF support.
Does AVIF take longer to encode than WebP?
Yes; single-threaded AVIF encoding can run 2-5x slower than WebP, which matters most if you encode on every request rather than once at upload.
What AVIF quality setting keeps photos looking sharp?
Stay in the 60-70 quality range for photographs, since AVIF below roughly 30 starts smearing fine texture in a way that looks worse than a larger JPEG.
Which format should I pick for a hero image?
Pick AVIF with a WebP fallback and a JPEG <img> base for hero images, since smaller bytes there directly speed up Largest Contentful Paint.
Image credits
- Cover — Photographer editing photos on a laptop with a DSLR and tablet, photo by cottonbro studio on Pexels (converted to WebP).
- Format comparison, file-size chart, and artifact zoom — generated by the author from a macaw-feather photograph (Pexels #36720663, photo by Kaca Skok) to show real measured JPEG / WebP / AVIF output at matched quality.
Use the free tools while you follow the guide.
Keep reading

2026-07-18
Convert HEIC to JPG: Free Tools and Batch Methods Compared
Convert HEIC iPhone photos to JPG with free web, Mac, Windows, iPhone, and command-line tools, compared by speed, batch support, and where each fits.

2026-07-18
JPG to PDF Guide: Combine Multiple Images Into One Document
Turn a set of JPG images into a single ordered PDF for portfolios, reports, and submissions. Free methods for each device, page ordering, and quality settings.

2026-06-28
How to Check Image Size, Dimensions, and File Format
Find an image's exact pixel dimensions, file size in KB or MB, DPI, and format. Methods for Windows, Mac, iPhone, Android, and online, plus what each value means.