Brave for Mac and the extensions you actually need at work

Most extension recommendations for Brave for Mac were written for a browser that no longer exists in that form. The extension format underneath the Chrome Web Store was replaced between 2025 and 2026, taking a portion of the recommended list with it, and Brave ships several of the remaining recommendations as built-in features. Follow an old list and the result is a browser carrying four extensions that duplicate something already running, two that were removed from the store, and one that quietly reads every page. The useful question is not which extensions to add. It is which ones are no longer needed.

The format changed and old guides did not

Brave installs extensions from the Chrome Web Store. The Chromium base is shared, so a listed extension generally works. The store itself changed underneath, though.

Manifest V2, the older extension format, was retired in stages. The official timeline states the endpoint plainly:

31 August 2026: All remaining Manifest V2 extensions removed from the Chrome Web Store Source: developer.chrome.com

Before that, on 24 July 2025, the format was turned off across all Chrome channels. So an extension that cannot be found by name today is usually not a search problem. It never migrated.

Brave made a partial exception. A small set of blocking and scripting extensions are kept working in Brave even after upstream Chromium drops support: AdGuard AdBlocker, NoScript, uBlock Origin, and uMatrix are the four named. The company is direct about the limits of that commitment, describing it as best effort and noting that support may be withdrawn if an extension goes stale. As the basis for a long-running work process, that is a conditional promise rather than a guarantee.

The wider point is that this was never only about ad blockers. Anything that inspected traffic or rewrote page content was affected: tab managers, tools that adjusted an internal business system's interface, small utilities built in-house. If a team distributed an extension a few years ago, checking whether it still loads is worth doing before it stops on a Monday morning.

Installing on a Mac, and what carries over

The install itself is unremarkable: a download from the official site, the usual drag into Applications, and a first-run dialog offering to bring everything across from another browser. That dialog is where the extension problem starts, because extensions are included in the import by default and taking the default reproduces the old browser exactly.

A more deliberate sequence works better. Import bookmarks and saved passwords, decline extensions, then add extensions back individually as a specific task demands one. Most setups finish the first week with four or five rather than forty, and the ones never re-added turn out not to have been missed. There is a diagnostic benefit as well: for that first week the browser has no third-party code running on pages, so anything that misbehaves can be attributed to the site or the browser without an afternoon spent disabling things one at a time.

History deserves a separate decision. It makes an article from last Tuesday findable, and it also makes a personal URL autocomplete in the address bar during a screen share. On a machine used for both work and personal browsing, starting with an empty history is a defensible choice.

One setting is worth deciding before the extension question at all: whether Brave becomes the system default browser. Left unset, links clicked in Mail or a chat app keep opening in the old browser, producing a state where the same service is signed in under different accounts in two places. During a trial, keeping the old default and opening Brave by hand keeps that from happening.

What Brave already does

Checking the built-in feature list before searching the store removes most of the work. Brave ships with several things that other browsers leave to extensions:

• Ad and tracker blocking, compiled into the browser
• Page translation
• A reader mode that strips a page down for reading
• Sync for bookmarks, passwords, and open tabs across devices
• A playlist feature for collecting media from pages

On the blocking specifically, Brave's position is that the extension is redundant:

Since Shields are patched directly onto the open-source Chromium codebase, they don't rely on MV2 or MV3. Source: brave.com

Stacking a second blocker on top is not additive. Two sets of rules applied to the same page can strip elements that were supposed to stay, which is a common cause of a form button that does nothing or an internal report that renders empty. When something breaks in a browser with a redundant blocker installed, isolating the cause costs an hour that did not need to be spent.

What is genuinely missing

Five categories survive the built-in list and are worth considering.

Need	Why built-in features fall short
Password management	Browser storage does not cover credentials shared with a team or used across other browsers
Screen capture and annotation	Full-page scrolling capture and markup are not native to macOS screenshots
Writing and translation assistance	Built-in translation converts a page. It does not apply a glossary or compare renderings
Time tracking	Logging billable time at the point of work reduces missed entries
Tab management	Partially addressable, with a ceiling discussed below

Two of these deserve a note. Password management is the one category where the browser's own feature is genuinely adequate for solo work and genuinely inadequate the moment a credential has to be shared, so the decision hinges on team size rather than on features. Screen capture is the opposite: macOS covers the common cases well enough that an extension is only justified by full-page scrolling capture or repeated annotation work.

Everything else that appears on a typical list, blocking, translation, reader mode, sync, is already present. Importing a list written for another browser without checking for this overlap is how an extension count climbs without anything getting better.

The deciding question for each candidate is how often the task occurs. Something used once a month does not justify a permanently resident extension, and the equivalent web service handles it at no ongoing cost.

Four checks before installing anything

An extension can read and modify the contents of the pages it runs on. For a browser used for client work, four checks are worth the two minutes.

Permission scope. An extension requesting access to read and change data on all sites can reach an admin console, a CRM, and an internal dashboard. Match the scope to the job, and where the extension supports it, restrict it to specific sites after installing.

Last update date. The store listing shows it. Continued updates through the format transition are a reasonable proxy for whether it will still work next year.

Publisher. Similar extensions with near-identical names are common, and ownership sometimes transfers after the original author steps away. Cross-check the listed publisher against the user count.

Overlap. Anything duplicating a built-in feature belongs on the reject list, for the reasons above.

Post-install maintenance has a fixed location too. The extensions page in settings opens the detail view for each one, where permissions and the sites it runs on can be narrowed later. That narrowing is easier a few months in than on day one, because by then it is clear where the extension is actually used.

Organisations sometimes restrict what may be installed. Checking policy before building a workflow around an extension avoids removing it all again later.

One habit is worth adopting alongside these checks: write down why each extension was installed, in a note kept with the rest of the setup documentation. Six months later the list contains at least one item nobody can account for, and without a reason recorded the safe choice is always to leave it running. A single line per extension turns the quarterly review from a judgement call into a short read. It also makes handover to a colleague possible, which matters more than it sounds, because an inherited browser full of unexplained extensions usually gets wiped and rebuilt from scratch.

Every extension runs on every page

Extensions cost something proportional to tab count, not to install count. Most work by injecting their logic into every page that opens, so thirty open tabs means that logic runs thirty times.

Brave, like other Chromium browsers, includes a task manager reachable from the Window menu. It lists memory and processor use per tab and per extension. When the browser feels slow, it settles in about ten seconds whether the cause is a page or something installed.

Opening it usually produces one surprise: a couple of always-on extensions accounting for a visible share of the total, often tools used a few times a year running on every page for all of it. Reducing the count does more than any setting. A quarterly pass is enough. For anything uncertain, disable rather than remove, and delete it a week later if nothing was missed, which preserves its configuration in the meantime.

Why tab managers disappoint

Tab management is the most requested extension category and the least satisfying one, and the reason is structural rather than a matter of quality.

An extension can only act inside the browser. Grouping tabs, naming groups, stashing and restoring sessions: all of it works, within that boundary. But macOS organises windows, not tabs. Mission Control, the app switcher, and the Dock all operate on windows and applications. Web apps living in the tab strip are invisible to every one of them, so an extension's organisation scheme becomes a second system to search, layered on top of the one the operating system already provides.

There is a second limit. When the same service has to be open under two accounts, an extension cannot help at all, because the requirement is separate storage rather than tidier display. In Brave that means containers or profiles, and no extension participates.

A third consideration applies to session-stashing extensions specifically: the stash lives inside the extension. If it stops being maintained, whether the accumulated list can be exported matters a great deal, and some of the extensions lost in the format transition were exactly this type.

Worth separating from all of this is the case of a handful of pages that simply need to stay open and visible. That is not a tab management problem and no extension is the right shape for it. Pinning helps a little, a second browser window helps more, and giving those pages their own windows helps most, which is the direction the rest of this points toward.

Taken together, the boundary is clear enough. Extensions are good at assisting work that happens inside a page. They are not able to fix where the pages live.

What to change first

Open the extensions page, remove anything that duplicates a built-in feature, and narrow the site permissions on what remains. Then count the web apps opened more than ten times a day: that count is the real problem, and no extension addresses it, because the fix is giving those apps windows the operating system can see. That arrangement is what Workspaces and Features describe, Supported apps lists which services are covered, and SpaceDeck shows the cost.