what is headless browser?
a real browser engine run without a visible window and driven by code, so pages render and javascript executes normally.
chrome and firefox both ship headless modes, usually driven through playwright, puppeteer or selenium. you get the whole stack: javascript execution, cookies and storage, service workers, a genuine browser tls handshake, a real dom to query after the page settles. for sites that build their content client side, or that gate access behind a scored javascript challenge, it is the only approach that works. the cost is resources. each instance holds a few hundred megabytes of ram and real cpu, and a single page pulls every image, font, video and third party tracker, so bandwidth per page runs ten to fifty times a plain http request. on metered proxy bandwidth that dominates the bill, which is why blocking images, media and fonts at the request level is standard practice. headless is also detectable by default through webdriver flags, missing plugins and gpu quirks; stealth patches reduce that, and running headed under a virtual display avoids some of it entirely.