mantrakp04/ardia

Ardia is worth a look if you like the sidebar-first browsing model but do not want to move to Arc or Dia. It is a macOS-only browser built on Chromium 155 with its own UI shell, distributed as a notarized DMG, and it asks you to accept a young project's rough edges rather than a finished product.

The problem

Sidebar browsers change where things live in a way that is hard to go back from: tabs, pinboards and the address bar move to one side, so the page gets the full window height. The practical problem is that the two reference points, Arc and Dia, are closed products on their own update schedules, and on macOS the alternative is usually a normal Chrome window with a vertical tab extension that does not feel like part of the app. Ardia aims at that gap: a Chromium engine with a native-feeling sidebar shell, tabs dragged into split panes, a floating video window when you switch away from a playing tab, and Chrome Web Store extensions with an "Add to Ardia" button. The result is meant to be the browsing layout you wanted without giving up Chrome's rendering, extensions or DevTools.

What's different

The meaningful difference is that this is not Electron or WebKit. Each window is Chrome's own framework from a patched Chromium build. That is why extensions, sites and DevTools should behave as they do in Chrome, and why the project can strip Google's background services and still claim the usual Chrome compatibility story. The README is unusually explicit about the de-Googled build, the H.264/AAC codecs Chromium leaves out, opt-in crash reports and end-to-end encrypted sync through your own iCloud Drive. There is also a detailed write-up of reverse-engineering one of Dia's New Tab Metal shaders, which shows how much of the visual polish is reconstructed rather than licensed.

The trade-off is the shape of the project. Building against a prebuilt engine means the UI and the engine have to stay in version lockstep; the README says the build downloads an engine matching your checkout (about 110 MB) and warns that when main is newer than the last release you should work from a release tag or fetch the nearest prebuilt. That is a normal cost for a browser, but it is a real one for anyone who wants to tinker. The UI shell is Expo plus React Native macOS, so the app is a hybrid native/JS project rather than a conventional Xcode-only codebase.

Who should try it

Try it if you are on an Apple Silicon Mac running macOS 14 or later, you liked Arc's sidebar and pinboard model, and you want a free GPL-licensed alternative that keeps Chrome extensions and DevTools. It is also a reasonable fit for someone who wants a Chrome-compatible browser with ad and tracker blocking included rather than installed from the Web Store as an afterthought.

It is a poor fit if you need Intel Macs, Windows or Linux, or if you depend on passkeys across devices: the README marks passkeys as partial, with Touch ID working but iCloud Keychain waiting on Apple. It is also a poor fit if you need a settled product with a long track record; this is a young codebase with a nightly channel and a release cadence you should check before committing.

Requirements and limitations

The hard requirements are narrow: macOS 14 Sonoma or later, Apple Silicon only, and the current build is 0.2.40. If you want to build it yourself, you need Xcode 26+, CocoaPods, Node 22.13+ and pnpm 11, and the engine download is Apple Silicon only. That already rules out much of the audience for a browser, and it means you should not expect a Linux or Windows port on any predictable schedule.

The ongoing limitation is the Chromium upgrade treadmill. Because the engine is a patched Chromium 155 and the shell is a separate layer, a Chromium update can require a rebuilt engine, which is why releases whose engine changed publish one. Translation is listed as on-device and macOS 26 only, and sync relies on your own iCloud Drive rather than a hosted account. The README also points to a nightly channel for fixes that have not reached a release, which tells you how the project expects users to balance stability and freshness. Check the release notes and the releases list before relying on a feature.

HotRepo's take

HotRepo first recorded the repository at 779 stars on 2026-10-25, with +102 stars that day and one appearance on the daily board; the repository was 127 days old at that point. That single dated snapshot is not enough to judge whether interest persists, so treat the popularity as unproven rather than as momentum. The concrete thing to verify before adopting is whether the engine/release pairing on the version you download matches the features you need, especially passkeys and sync, since both are documented as partial or dependent on Apple's side.