Optimize your web assets by converting JPG files to the highly efficient AVIF format.
Not every JPEG benefits equally from AVIF conversion. Run through this checklist before you commit to the switch:
The numbers are not marketing claims. Independent testing by the AVIF working group and Netflix (which helped develop the format) shows consistent results:
The trade-off is encoding speed. JPEG encodes almost instantly; AVIF encoding uses the AOM AV1 codec, which is computationally heavier. For a batch of 500 product photos, expect encoding to take 3–8 minutes on a standard laptop — a one-time cost for permanent bandwidth savings.
Most converters present a "quality" slider, but AVIF encoding involves several interacting parameters. Here's what each one does in practice:
Quality (quantizer): This is the primary knob. AVIF uses a scale of 0–63 where 0 is lossless and 63 is the most aggressive compression. Unlike JPEG's 0–100 scale, lower numbers mean higher quality. For typical web use, 30–45 is the sweet spot.
Speed: Controls the encoder's effort level (0–10). Speed 0 runs the most exhaustive compression search and produces the smallest files. Speed 10 encodes fastest but leaves compression opportunities on the table. For one-off conversions, use speed 4–6. For batch processing, bump it to 8.
Chroma subsampling: JPEG uses 4:2:0 chroma subsampling by default, which discards color detail. AVIF can encode at 4:4:4, preserving full color resolution. This matters for images with fine colored lines or saturated gradients — think infographics, charts, and colorful product photos.
Alpha channel: JPEG doesn't support transparency, so this setting is irrelevant when converting from JPEG. It becomes relevant when converting from PNG, where you'd want to preserve the alpha.
If you're migrating a media library — say, 2,000 product images from a JPEG-based CMS to AVIF — the conversion itself is the easy part. The hard parts are the edges cases that surface only at scale:
Mixed color profiles: A batch of images collected over years will contain sRGB, Display P3, and occasionally untagged files. Untagged files default to sRGB in most viewers, but some AVIF encoders treat them as linear RGB, producing washed-out output. Check the first 10–20 files manually before processing the full batch.
Orientation metadata: JPEG files from cameras contain EXIF rotation tags that viewers interpret on display. When you convert to AVIF, some encoders strip this metadata, causing images to appear rotated. Use an encoder that preserves or bakes in the EXIF orientation.
Progressive JPEG edge case: Progressive JPEGs load in multiple passes. Some AVIF conversion tools decode only the first scan, producing lower-quality output. Verify that your converter fully decodes the JPEG before encoding to AVIF.
For smaller batches — say, a few dozen images for a blog redesign — Pixes.app handles the entire job client-side with no file size limit and no upload queue. Drop the files, set your quality, and download the results.
AVIF is not universally superior. There are specific scenarios where converting JPEG to AVIF adds complexity without meaningful benefit:
Email attachments: Most email clients do not render AVIF images inline. The recipient will see a broken image icon or a file attachment they can't open without downloading. Stick with JPEG or PNG for email.
Print workflows: Print services expect JPEG, TIFF, or PDF. Submitting AVIF files will either be rejected or trigger a support ticket.
Already-compressed JPEGs below 60 KB: These files have already sacrificed significant detail. Re-encoding into AVIF at any quality setting will either maintain the same size or increase it, with no visible improvement.
CMS platforms without AVIF support: WordPress added AVIF support in version 5.8, but many hosted platforms and legacy CMS installations still lack it. Check your platform before converting your entire library.
If you're serving AVIF to modern browsers with a JPEG fallback for older ones, the JPEG version should also be as small as possible. Running your originals through a JPEG compressor strips unnecessary metadata and applies lossless optimization — typically reducing file size by 8–15% without any visible quality change.
This two-pronged approach — optimized AVIF as primary, optimized JPEG as fallback — is what large-scale sites like Shopify and Cloudflare serve. The AVIF handles 80%+ of traffic (anyone on a modern browser), while the compressed JPEG covers the remaining legacy visitors.
Combined savings: if your original JPEG was 600 KB, the AVIF version might be 280 KB and the fallback JPEG might be 510 KB. Multiply that across a 200-image product catalog and you're looking at tens of megabytes saved per page load.