Optimage Alternatives for Mac: Intel-Only Since 2021

Optimage is a better answer than we are in several real situations, and pretending otherwise would be silly when the download is one click away and you can test it yourself.

You are on an Intel Mac. Rosetta is irrelevant to you. Optimage runs natively, it is $15 against our $29, and the free tier gives you 24 images a day forever if that is all you need. Nothing about the 2021 release date stops it from compressing a JPEG correctly in 2026, because image formats did not change underneath it.

Your files are SVG, ICO, or ICNS. Smol does not touch any of those. If your job is minifying vector graphics out of Illustrator or building favicons, Optimage covers ground we do not.

You want a tool with no quality slider. Optimage’s design position is that you should not have to pick a number: it aims for visually lossless and decides the encoder settings for you. Some people strongly prefer that. Smol will do the same thing on a preset, but it also hands you quality 0 to 100 and a target-KB field, which is either freedom or homework depending on your temperament.

The thing to weigh against all of that is not quality. It is the Rosetta clock, and that clock only runs on Apple silicon.

26 November 2021, on three independent pieces of evidence that all agree.

The first is the app’s own update feed. Optimage uses Sparkle, the standard macOS updater, and its appcast lives at optimage.app/appcast.xml. Fetched on 26 September 2026, it contains a single <item>: version 3.5.1, build 195, with <pubDate>Fri, 26 Nov 2021 21:25:19 +0300</pubDate>. That feed is what every installed copy of Optimage checks when it looks for updates. It has had nothing new to say for close to five years.

The second is the published version history page. Its newest entry is dated 26 November 2021 and reads, in full, “Fixed HEIF detection”. Below it sits 24 November 2021, the last substantial release, which added multithreaded video compression and APNG and HEVC conversion.

The third is the binary itself. We downloaded optimage-mac.zip on 25 September 2026 and unpacked it. The bundle’s Info.plist reports CFBundleShortVersionString 3.5.1 and CFBundleVersion 195, matching the appcast exactly, and the unpacked directory carries a modification date of 26 November 2021.

Five years without a release is not automatically a problem. JPEG did not change. PNG did not change. An encoder that produced good output in 2021 produces the same output today, and Optimage’s remains good. What did change is everything around it. AVIF became the default choice for web images, and Optimage does not list it as an output. And Apple put a deadline on x86_64 binaries.

Only through Rosetta, and only for one more macOS release.

We checked this directly rather than inferring it. After downloading the current build we walked every executable file in Optimage.app and ran lipo -archs on each one. Thirteen Mach-O binaries came back, and every single one reports x86_64 alone: the main application, the bundled Sparkle updater and its two helper tools, and the nine command line tools in Contents/MacOS/cli, including ffmpeg, Ghostscript, and cwebp. There is no arm64 slice anywhere in the bundle. This is not a universal binary that merely lacks optimization. It is an Intel application.

On an M-series Mac, macOS translates it with Rosetta 2, which works and has worked well for six years. The relevant change is that Apple has now put an end date on it. In a developer note published on 1 September 2026, Apple wrote that on “macOS 26.4 or later” users “may receive a system notification when launching apps that rely on Rosetta, alerting them to update to an Apple silicon native version”, and that macOS 27 is the “final release to support Rosetta”, after which “Intel-only apps will no longer run on Mac computers with Apple silicon”.

macOS 27 shipped this month. So the practical position for anyone on an M1 or later is: Optimage works today, it will show you a nag, and the next major macOS version is the one where it stops launching. The update feed shows no sign of a native build in progress.

Smol, by contrast, ships separate native downloads for Apple Silicon and 64-bit Intel. Squoosh has no architecture to worry about at all, because it runs in a browser.

The method is better than anything else in this category. The results are from 2019.

Credit first, because it is deserved. Optimage’s benchmark page does something almost no compression tool does: it names its corpora (the PNG test corpus by Cédric Louvrier and the Guetzli photo dataset by Jyrki Alakuijala, both publicly downloadable), it states its scoring rule (a flip test on a calibrated MacBook Pro Retina display, one point if an image is indistinguishable from its original and zero if it is not), and it reports Butteraugli distances alongside the human scores so you can see whether the two agree. Its headline numbers: a total human-rated score of 100% for Optimage against 81% for Guetzli, 58% for JPEGmini, 48% for TinyPNG, 44% for ImageOptim, and 13% for Google Photos. Most competitors publish a percentage and a hope.

Two caveats a careful reader should hold. The homepage describes these as “third-party tests”, but the link goes to Optimage’s own benchmark page, so they are first-party results run by the vendor. And the page is old: a conditional request to optimage.app/benchmark returns Last-Modified: Fri, 20 Sep 2019 08:48:23 GMT. Those ImageOptim, TinyPNG, and JPEGmini scores describe the versions of those tools that existed in September 2019. Encoders moved in the years since. The benchmark is a snapshot, not a scoreboard.

One number of our own, measured rather than sourced, using Optimage’s own sample gallery so it is reproducible. We downloaded the nine sample images published on optimage.app (five JPEG, four PNG, 2,068,780 bytes in total) and compressed them as a single batch with Smol: output format original, image quality 75, metadata stripping on, on an Apple M2 Pro with 16 GB running macOS 27.0 (build 26A428). Result: 559,774 bytes out, a 72.9% reduction at unchanged pixel dimensions, in 4,338 ms for the whole folder. Optimage publishes its own figures for those same nine files on its homepage, totalling 797,259 bytes.

Do not read those two totals as a head-to-head. They are not quality-matched: quality 75 trades visual fidelity for bytes, while Optimage’s default aims for visually lossless, which is a harder target and a larger file by design. A fair comparison would need the flip test Optimage itself describes, and we have not run it. What the measurement establishes is the method and the machine, so you can repeat it.

Four options, each genuinely the right answer for a different person.

The honest summary of that table: if Optimage already sits in your dock on an Intel Mac and does the job, keep it. If you have moved to Apple silicon, you are choosing between a free native tool and a paid one, and the deciding question is whether you also need AVIF, PDFs, video, or a file-size target.

Each option has its own page here, written the same way. ImageOptim alternatives for Mac covers the free lossless route. Squoosh alternatives documents the other dormant favorite in this category, where the maintainer has said on the record that Google will not maintain it. Smol vs TinyPNG weighs a local app against a cloud service, and compressing images on Mac, every method compared covers Preview and sips as well.

If the native route is the one you want, Smol is a one-time $29 purchase for one Mac, with no subscription and no per-file limits.