Does Shopify compress images? Yes, when it serve...

Does Shopify compress images? Yes, when it serves them: Chrome got a median 11% of the source, 25% at full size

By SJ
7 min read
Blue and violet gradient cover with the white text Does Shopify compress images? Chrome got a median 11% of the source, a field of faint squares on the right and the Iminify mark.
On this page

Yes, in the files it serves, and not by one fixed amount. On 9 November 2026 we loaded 78 pages from 43 public Shopify stores in Chrome 156 and asked Shopify's image servers for 459 of the images on them. The file Chrome got was a median 11% of the size Shopify reports for the source. At full size, the source's own pixel size instead of the width the page asked for, it was 25%.

Shopify's help says images "aren't compressed when you upload them to the Shopify admin". Its CDN page says "File size is automatically optimized when you upload to Shopify". We uploaded nothing, so this is about what Shopify sends, not what it keeps.

Shopify's pages disagree, and so do the guides

Shopify's help page on uploading images adds that served images "might undergo compression for high-quality images". The page for its CDN lists "File compression" as something that happens on upload. Four guides we read on 9 November split three ways. Fudge and Hyperspeed say Shopify doesn't compress source files, imagecompressor.com that its resizing "doesn't aggressively recompress the image data", and HulkApps that it "intelligently compresses" images and you can't switch that off. None shows a measurement.

Chrome's file was a median 11% of the source Shopify reports

Every response we fetched from Shopify's image servers carried a source-length header, the size of the file it was made from. On the sample photo its CDN page offers, a 1,850 by 1,233 JPEG, it reads 368,109 bytes: the page's "Original file 359.48 KB", counted in 1,024s.

From each store we took the six addresses that asked for the largest width (200 pixels or more) and up to six more at random, for 459 images from 39 stores. We requested each address three ways: as Chrome does, as a browser with WebP but no AVIF or JPEG XL does, and as a browser without WebP does. The first and last are the pair our Squarespace post compared. The table leaves out one GIF, and both copies are at the width the page asked for:

Source file Images Chrome's file, share of the source Browser without WebP, share of the source
JPEG 289 10% (quartiles 4% and 28%) 16%
PNG 131 9% (2% and 21%) 72%
WebP 38 61% (20% and 88%) 123%

Together the 459 sources weighed 681 MB and Chrome's copies 42 MB. Part of that is resizing. The 388 addresses that name a width asked for a median 948 pixels. At the bare address, the source's own pixel size (the same as a request for the 5,760 pixels that Shopify's image_url filter allows at most, on 457 of 459 images), Chrome's file was a median 25% of the source, with quartiles at 11% and 71%. The biggest source was 14.6 MB.

Chrome gets WebP for big images, JPEG XL and AVIF for small ones

Chrome 156's Accept line starts with image/jxl,image/avif,image/webp. Shopify's pages name WebP and AVIF, and none of the four Shopify pages we saved mentions JPEG XL. Across 3,663 image addresses on 39 stores, Chrome was sent WebP for 59%, JPEG XL for 25%, AVIF for 12%, PNG for 2% and JPEG for 2%. For the 2,586 addresses that name a width, the format follows it:

Width in the address Addresses JPEG XL AVIF WebP JPEG or PNG
under 300 659 71% 26% 2% 1%
300 to 799 1,168 17% 10% 70% 3%
800 to 1,399 470 0% 0% 99% 1%
1,400 and over 289 0% 0% 97% 3%

JPEG XL is a quarter of the addresses and 5.6% of the bytes: its median address asked for 80 pixels. Shopify's pick isn't always the smallest. Of the 459 images, 66 came back to Chrome as JPEG XL and 33 as AVIF. A WebP-only browser got a WebP for 47 of the 66 and all 33, and the JPEG XL file was larger than that WebP in 9 cases and the AVIF in 8.

The default is JPEG at 85 and WebP at 85 to 89

Shopify's documentation lets a theme set quality from 10 to 90, and says that leaving it out "isn't equivalent to setting it to 100". We asked 48 images, once as Chrome and once as a browser without WebP, for every quality from 60 to 90 and looked for the one that gave a file the size of the default. The browser without WebP got a JPEG for 36 of them, and 85 matched all 36. Chrome got a WebP for 40, every one matched, and the number fell as the copy grew:

Size of the WebP Default quality Files
0.10 to 0.23 megapixels 89 6
0.36 88 2
0.63 to 0.71 87 6
0.81 to 1.35 86 13
1.46 to 4.77 85 13

JPEG XL matched 90 on 6 of 6. These are values of Shopify's own quality parameter, not another tool's scale, and JPEG 80 is not WebP 80 shows why they can't be assumed equal.

Only PNG sources give a lossless reference: the PNG a browser without WebP gets, truecolour (with or without transparency) for 117 of the 131 and a palette for 14. Chrome was sent a WebP for 93 of the 131, and the 92 that scored had a median of 81.4 on SSIMULACRA 2, with quartiles at 76.2 and 83.8. Of the 92, 52 scored 80 or more, where the metric's authors say a distortion isn't noticeable side by side. For JPEG sources the only reference is Shopify's own JPEG, and the WebP scored a median 75.5 against it, which compares two lossy copies. We didn't measure what Shopify's re-encode costs a JPEG source in quality.

Some copies match the source, and four WebPs carried a make and model

At the source's own size, the copy for a browser without WebP was within 60 bytes of the source for 127 of the 459 images, so Shopify doesn't send a re-encode for every image. Chrome's copy was that close for 27. At the width the page asked for, none of Chrome's files carried GPS coordinates. Four of them, WebPs on two stores, carried an EXIF make and model, three of them Canon cameras. The copies for browsers without WebP, at the source's own size, carried none.

What to upload

  1. A file inside Shopify's limits. For product and collection images its help says 5000 by 5000 pixels or 25 megapixels and under 20 MB. For other uploads its other page says 20 megapixels and 20 MB, counting a megabyte as a million bytes (both read 9 November 2026). If a file is over, resize it with Iminify first.
  2. Don't compress it first for Chrome's sake. Shopify says images "don't need to be manually compressed prior to uploading", and Chrome gets a re-encode anyway, most often a WebP. We couldn't test what a lossy first pass adds, because that needs an upload. Through Next.js and wsrv.nl, only a lossless first pass left the result unchanged, and Iminify's Lossless level does that to a JPEG.
  3. JPEG for photographs, if you have the choice. Shopify's two help pages differ: one lists PNG first for product images, the other says to use JPEG for them. For each image, a browser without WebP got a median 1.4 times Chrome's file for JPEG sources and 6.8 times for PNG sources. Such browsers are few: caniuse counts WebP support in 97.2% of browser use.

How we measured

On 9 November 2026 we loaded the home page and a product page of 35 stores, and the home page alone of 8 where we found no product in the sitemap, in Chrome 156.0.8078.81, headless at 1440 by 900 pixels and device scale factor 1. The stores were 50 well-known brands we picked: 43 loaded and ran on Shopify, 3 loaded and didn't, and 4 refused the request. That makes a convenience sample, on desktop, where a phone asks for smaller widths. "Shopify's image service" means addresses with /cdn/shop/ in the path or on cdn.shopify.com/s/files/. Four stores loaded none, leaving 39 with 3,663 addresses. MB are 1,000,000 bytes.

The three requests were the address Chrome used with Chrome 156's Accept line, with image/webp,image/*,*/*;q=0.8 and with image/*,*/*;q=0.8, and the first and last again with the width, height, crop, format and quality removed. A copy Shopify cached months ago can differ from a fresh one: asked again with a changed v parameter, 1,133 of 1,377 responses gave the same bytes, and the 244 that didn't had been cached a median 48 days, against 7 for the rest. The JPEG XL and AVIF comparisons and the quality matches use the fresh copies, and the share of the source is the same 11% either way. Scores are libjxl 0.12.0's SSIMULACRA 2, on the first copies we fetched, of Chrome's WebP (decoded with dwebp 1.5.0) against the copy for a browser without WebP at the same address.

Questions this post answers

Share this post

Share on X
Share on Facebook
Share on LinkedIn
Share on Reddit
Share on Hacker News
Email
Copy link
Link copied
The same photo of a canoeist twice: upright on the left under the label "tag read", and lying on its side on the right under the label "tag gone".

Why does my photo upload sideways? The rotation lives in a tag, and stripping metadata throws it away

We stored a photo on its side with the EXIF tag Orientation 6 and opened it in Chrome 156. It stood upright. Then exiftool -all= removed the metadata, and the photo lay on its side. So did the copies saved by magick -strip and jpegtran -copy none. That is how a photo can look right on your phone and arrive sideways on a website. The file holds the pixels the way the camera sensor wrote them, plus a tag that tells a viewer how to turn them. A site that ignores the tag, or software in its pipeline that drops it, shows the pixels as stored. The fix is to turn the pixels themselves so no tag is needed, and the steps are below. Removing metadata doesn't do that.

7 min read
Three crops of the smallest text on the Inter sheet with the fields Gemini read: the PNG (228 KB) 32 of 32, JPEG quality 40 (172 KB) 32 of 32, JPEG quality 5 (76 KB) 9 to 14 of 32, where the digits are visibly smeared.

Does compressing an image change what Gemini reads? Full-colour JPEG quality 95 to 40 didn't, shrinking to 600 pixels did

Gemini 3.8 Flash read the same number of fields from a full-colour jpegli JPEG saved at quality 40 as from one saved at quality 95, on both of the text sheets we tested. A copy shrunk to 600 pixels wide, a PNG with no JPEG loss in it, read 15 fields fewer than the quality 95 file on both, comparing medians. A field is one code, price, date or 12-digit number, and each sheet holds 192 of them as random strings, so there's nothing to guess from. Anthropic's vision page says "heavy JPEG compression can make text difficult to read" and gives no number. Google's image page advises "Use clear, non-blurry images" and doesn't mention compression. We read both pages on 8 November 2026. These numbers are for Gemini 3.8 Flash, through the API, on 8 November 2026.

7 min read
Bar chart of remove.bg's sample cutout in bytes: PNG 6,731,926, Smart PNG 2,407,602, Smart WebP 356,282, Smart AVIF 194,550

Background remover PNG too big? One cutout went from almost 7 MB to 356 KB as WebP

The sample cutout that remove.bg publishes on its API page is a 6,731,926-byte PNG. Through Iminify at Smart, its default lossy level, the same file came back as a 356,282-byte WebP (356 KB) with every transparency value unchanged. The AVIF was 194,550 bytes and changed transparency values by up to 29 of 255; the PNG was 2,407,602 bytes and kept 32 of the 256 values. So save the cutout as WebP, which changed no transparency value, and keep the PNG from the remover as your master. PNG to WebP does it. Below is what each choice gives up.

6 min read

Try it on your own images

Drop a photograph in, pick Smart, and compare the result with the original before you download. Free, no account needed.