An Arc browser review now that development has stopped

Most people reading an Arc browser review today are not deciding whether to install it. They already have it, they have spent a year arranging Spaces and pinned tabs to fit how they work, and they want to know whether that arrangement is safe to keep building on. That is a different question from "is this a good browser", and it deserves a different answer.

The short version: Arc still runs, still downloads, and still receives the underlying Chromium updates. New feature work on it has stopped, and the vendor has said so in public. Whether that matters depends almost entirely on which parts of Arc you actually use, and the vendor has published numbers suggesting that for most people the answer is fewer parts than they think.

What the vendor has confirmed about Arc's future

In May 2025, The Browser Company published a long letter to Arc members explaining why it had pivoted to a second product, Dia. The letter is unusually direct about what happened to Arc. It states that the company kept doing regular Chromium upgrades and security fixes, but stopped building the core product experience. It also states plainly that shutting Arc down was not the plan.

The same message now sits on the Arc homepage itself, in a single line above the download buttons:

Arc receives Chromium updates only. For active security patches and enterprise-grade protection, download Dia instead.

Source: arc.net

Two things follow from that sentence, and they point in opposite directions. Chromium updates are not nothing. They carry the bulk of the browser engine's security fixes, which is the part that matters most for something you type passwords into. But the vendor is also drawing a distinction between those updates and "active security patches", which suggests the browser shell that sits on top of Chromium, the part that makes Arc feel like Arc, is not getting the same attention.

The letter also closes off the outcome many users hoped for. Arc is not a plain Chromium fork. It runs on internal infrastructure the company calls ADK, which is also the foundation of Dia, so releasing Arc as open source would mean releasing the thing the company still trades on. The letter says this directly and does not promise a date.

What that means in practice

A browser in this state is usually fine for a while and then, one day, is not. The failure mode is rarely dramatic. A site changes how it handles a login, an extension API shifts, macOS changes how windows behave after an upgrade, and the fix that would normally arrive in a point release does not arrive. Nobody can tell you when that day comes. What you can do is know in advance which of your habits would break, and how long it would take to rebuild them somewhere else.

Which Arc features people actually use

The most useful part of that letter, for anyone weighing a move, is the usage data the company chose to publish about its own product. It is rare for a vendor to print numbers this unflattering.

Only 5.52% of DAUs use more than one Space regularly. Only 4.17% use Live Folders (including GitHub Live Folders).

Source: browsercompany.substack.com

Read that twice if you are one of the people who built an elaborate Space layout. Roughly one Arc user in twenty regularly used more than one Space. The company's own framing is that Arc's numbers looked like a specialised professional tool rather than a mass market browser, and that the features enthusiasts talked about were used by almost nobody.

For a review, this cuts both ways. If you are in the 5.52%, you are using the product in the way it was designed for, you are getting real value from it, and you are also the user the vendor has the least commercial reason to keep serving. If you are not, then the thing keeping you in Arc is probably the sidebar, the command bar and the general calm of the interface, and those are much easier to replace.

The honest test

Open Arc and count. How many Spaces do you switch between in a normal week, not in theory but in fact. How many separate accounts are signed in across them. If the answer is one Space and one account, almost any modern browser will do, and this review can stop being interesting to you. If the answer is three Spaces holding five accounts across Gmail, Slack and a project tool, you are not really using Arc as a browser at all. You are using it as an app container that happens to render web pages, and that is a category with other members.

What Arc still does well

Fairness matters here, because "development stopped" reads as "abandoned" and that is not what is happening. Arc on a Mac today still gives you a vertical sidebar instead of a horizontal tab strip, Spaces with their own profiles, split view, a command bar, and an interface that a lot of people find genuinely pleasant. It installs on Apple Silicon and Intel, there is a Windows build, and there is an iPhone app.

None of that has been withdrawn. The browsing experience you liked in 2024 is the browsing experience you get in 2026. If your use is a single Space, a handful of pinned tabs and a preference for how the sidebar looks, the practical risk of staying is low and the cost of moving is mostly emotional.

The risk rises with how much of your working setup lives inside it. A browser holding six signed in accounts and a year of arranged pinned tabs is infrastructure, and infrastructure on a product with no feature roadmap is a position worth reviewing on purpose rather than by accident.

The cost of staying, stated plainly

Reviews of products in this state tend to end in vague advice about keeping an eye on things. That is not useful. The cost of staying on Arc can be written out.

There is no licence cost. Arc is free and there is no paid tier to be pushed onto, which removes the most common reason people leave a browser.

There is a compatibility cost that grows slowly. Every month that passes, the gap widens between what websites assume a browser does and what a browser with no active feature team was built to do. This shows up first in the places browsers change fastest, which right now means video calls, extension permissions, and anything touching a new web standard for identity or payments.

There is a support cost. When something breaks in a product with an active team, you file a report and wait. When something breaks in a product that is being maintained rather than developed, the honest answer is that a report about a non security issue may sit. Plan on solving your own problems.

And there is a concentration cost, which is the one people underrate. The more of your working life you put into any single tool, the more expensive it becomes to move, and the less freely you can decide. That cost applies to every browser, not just this one. It is simply higher when the tool is not being built any more.

How the alternatives actually differ

People leaving Arc tend to go in one of three directions, and they are not interchangeable.

The first is back to a mainstream browser. Chrome profiles or Safari profiles give you separated cookie jars and separate windows, with a vendor that is not going anywhere. You lose the sidebar arrangement and gain certainty.

The second is toward another AI focused browser, including Dia from the same company. This is the direction the vendor is pointing, and it is coherent if what you want next is a chat interface welded to your tabs.

The third is toward a browser built specifically to hold web apps rather than web pages. This is the category worth understanding if Spaces were the reason you were in Arc, because these products treat "one app, one isolated container, always running" as the default rather than a power feature. Prices and limits below are taken from each vendor's own pricing page as of September 2026.

Product	Platforms	Free plan	Paid
Arc	macOS, Windows, iPhone	Full browser, free	None
Wavebox	macOS, Windows, Linux	2 groups, 2 apps each, 1 extension	8.33 USD per month, billed yearly
Shift	macOS, Windows	5 Spaces, 10 apps each	199.99 USD per year
Rambox	macOS, Windows, Linux	Unlimited apps, 2 instances per app, workspaces are paid	7 USD per month, 70 USD per year
Ferdium	macOS, Windows, Linux	No limits, open source	None

The comparison with Wavebox and the comparison with Rambox go through the same ground feature by feature, including the places where each product is the stronger choice.

The thing to check before committing

Whichever direction you pick, check one item first: how many accounts of the same service you can be signed into at once, and whether that sits behind the paywall. This is the single feature most people in this category are actually shopping for, and it is the one most often placed on the paid tier. If two Gmail accounts open side by side is the reason you are moving, confirm it works on the free plan before you rebuild anything. The pricing page for this kind of tool is usually where the real limits are written down, not the feature list.

Moving a Space without losing a day

The part that keeps people in place is not preference, it is the expected cost of rebuilding. It is smaller than it looks if you do it in the right order.

Arc's Spaces are built on Chrome profiles underneath, so your bookmarks and passwords are exportable through the standard Chromium paths rather than trapped in a proprietary format. Export bookmarks per profile before you do anything else. Passwords belong in a password manager rather than in any browser, and if they are currently in Arc, that move is worth making regardless of what you decide about the browser itself. Google's own documentation covers exporting from a Chromium based browser.

Then rebuild one Space, not all of them. Pick the one you use most, set it up in the new tool, and run both browsers side by side for a week. This is the step people skip, and it is the one that tells you whether the new arrangement holds. If isolated workspaces behave the way you need, the rest is copying. If they do not, you have lost a week instead of a month.

What to change first

Count your Spaces and your signed in accounts this week, and write the number down. If it is one and one, stay on Arc and stop reading reviews about it. If it is three Spaces holding five or more accounts, your setup is an app container rather than a browser, and it should be running on something still being built. Rebuild one Space in a tool designed for that job and give it seven days before deciding anything else. SpaceDeck is one place to start that week.