Virtual browser
A browser that runs on remote infrastructure and is driven over the network, rented by the session instead of run on your own machines.
A virtual browser is a real browser running somewhere else, on a provider's infrastructure, that you control over the network as if it were local. Instead of installing and running headless browsers on your own machines, you connect to a hosted one, drive it through a standard automation protocol, and get back the rendered result. It packages the heavy, awkward part of browser-based scraping as a service you rent by the session.
The appeal is that running headless browsers well at scale is genuinely hard, and this offloads it. Each real browser instance eats hundreds of megabytes of memory and real CPU, so a fleet of them is an infrastructure project in itself, with all the scaling, crashing and resource management that implies. A virtual browser service absorbs that: it runs the fleet, keeps the browsers patched and plausible, and often layers in the proxy rotation and fingerprint management that browser scraping needs anyway, so you get a rendered page from one network call.
It overlaps with two neighbours and is worth distinguishing from both. A headless browser is the same technology run on your own machines, so a virtual browser is essentially that as a managed service. A scraper API is a step further: it hides the browser entirely and returns just data, where a virtual browser still gives you a browser to drive. The choice between them is how much control you want, full programmatic control of a browser you did not have to host, versus a simpler endpoint that decides the how for you.
The trade-offs are the usual managed-service ones. You pay a premium over self-hosting that is worth it up to the scale where running your own fleet becomes cheaper, and you inherit the provider's choices about browser configuration and behaviour. For teams that need real rendering but do not want to become browser-infrastructure operators, it is often the pragmatic middle, and for the multi-identity account work the antidetect entry describes, a hosted browser with managed fingerprints and a paired proxy can be the whole toolchain in one place.
Frequently asked questions
What is a virtual browser?
A real browser running on a provider's remote infrastructure that you drive over the network as if it were local. Instead of hosting headless browsers yourself, you connect to a hosted one, control it through a standard automation protocol, and receive the rendered result. It turns the heavy part of browser-based scraping into a service rented by the session.
What is the difference between a virtual browser and a scraper API?
How much they hide. A virtual browser gives you a full browser to drive remotely, so you keep programmatic control without hosting it. A scraper API goes further and hides the browser entirely, returning just the data for a URL. One is managed infrastructure you still operate; the other is an outcome you request. Choose by how much control versus simplicity you want.
Do I need proxies with a virtual browser?
Usually, and many virtual browser services bundle proxy rotation and fingerprint management precisely because browser scraping needs them. The rendered session still exits to the target from some address the site judges, so network trust matters exactly as it does with a local headless browser. Check whether proxies are included or whether you supply your own.
Back to the full glossary.