Guides, playbooks & product news

What we write about, and why

Most writing about antidetect browsers is either a feature list with a discount code attached or a warning that the whole category is disreputable. Neither helps somebody who has to make two accounts work on Monday. What is missing is the middle: how detection actually operates, which parts of it a browser can address, which parts it cannot, and what other people have found out the expensive way.

So the articles here cluster into four groups. How detection works -- canvas, WebGL, audio and font fingerprinting explained properly, what a checker result means, and why consistency between values matters far more than uniqueness of any one of them. Proxies -- residential against datacentre against mobile, how many you need, rotation strategies, and the mismatch between proxy country and browser timezone that undoes more setups than any detection method. Running accounts at scale -- what breaks first past a few hundred profiles, how teams share access without sharing passwords, and how sessions survive restarts and moving between machines. And automation -- driving browsers without announcing it, why the usual tooling is detectable by default, and what to do instead.

If you are new to the subject, three articles are worth reading before the rest, because almost everything else assumes them. Start with what browser fingerprinting actually is and which values carry the signal -- it explains why signing out and private windows do not separate accounts, which is the misunderstanding most expensive setups are built on. Then the proxy piece: the categories, what sharing an address really costs, and the timezone-against-exit mismatch that quietly undoes otherwise careful work. Then how antidetect browsers work, which covers the difference between changing values in the browser engine and injecting JavaScript over the top -- the distinction that determines where a tool breaks under a serious detector.

The archive runs deep -- several hundred articles at this point -- so the search engine is genuinely the better way in than the pager once you know what you are looking for. Every piece stands alone: none assumes you have read the others, beyond the three starting points named above, and each links its prerequisites where it has any. The pagination below is ordered newest first, which means the fundamentals sit in the middle pages rather than at either end -- another reason to start from the three named articles rather than from page one.

If you are looking for something specific and cannot find it, the community is usually faster than searching: somebody has generally hit the same question, and the thread will contain the parts an article would have left out. Requests for subjects we have not covered are welcome there too, and several of the pieces here started as exactly that.

Nothing here is generated. Each article is written by somebody who has run the setup being described, which is why the specific numbers keep reappearing -- roughly five concurrent profiles per 4 GB, a week as the minimum honest test, the timezone-against-exit check after every proxy change. Those are not editorial positions; they are what the work actually looks like, and an article that contradicted them would be describing something none of us has done.

A fair question is why a company selling an antidetect browser should be trusted to explain how detection works, given the obvious incentive to make the problem sound larger than it is. The check available to you is the one worth applying to any vendor’s writing: does it tell you when the answer is not to buy anything. Several of these articles conclude that a cheaper approach is sufficient, that a problem is behavioural rather than technical, or that no tool fixes the thing being asked about. Those are the pieces to read first, because they are the ones an advertisement would not contain.

A note on how these are written and maintained. Detection changes, so an article about it has a shelf life -- what a platform checked last year is not what it checks now, and advice that was sound when published can quietly become wrong without anything visibly breaking. Where a piece states something specific enough to go stale, it says when it was verified, and the ones covering fast-moving ground get revisited rather than left standing. If you find something here that no longer matches what you are seeing, the community is the fastest place to say so, and it is genuinely useful rather than merely welcome.

Two things you will not find. Guarantees that a platform will accept your accounts, because no vendor has access to that decision and pretending otherwise is the industry’s worst habit. And advice on defrauding anyone -- the tool is for holding accounts you are entitled to hold, and the writing follows the same line.