Browser Fingerprint Test

No. Cookies and fingerprinting are unrelated mechanisms. A cookie is a piece of data the site stores in your browser; a fingerprint is derived live, on each visit, from your hardware and software configuration — GPU, installed fonts, screen size, and similar signals. Clearing cookies removes stored identifiers but does not change your GPU model, your installed fonts, or your screen resolution, so the same fingerprint hash is reconstructed the next time you load the page.

Only partially. Private browsing modes prevent cookies and history from persisting after you close the window, but the underlying signals used for fingerprinting — canvas rendering, WebGL renderer string, hardware concurrency, installed fonts — are identical in a private window and a normal window. A site can still compute the same fingerprint hash in both modes; what changes is only whether a cookie-based identifier survives the session.

A VPN changes your IP address and can reduce IP-based geolocation, but it has no effect on the 14 signals this tool measures. A VPN does not alter your GPU, fonts, screen resolution, or audio stack, so your uniqueness score is expected to be the same with or without a VPN connected. If you want to reduce IP-based tracking specifically, pair this test with our WebRTC and DNS leak checks, which are the tests a VPN is actually meant to affect.

Somewhat counterintuitively, no — it can make you more identifiable. The DNT header is honored by very few sites and disabled by default in most browsers, so a request with DNT enabled is itself a distinguishing signal, since only a small minority of visitors send it. Research from the EFF's Cover Your Tracks project (formerly Panopticlick) documented this effect directly: turning on a rarely-used privacy signal can increase entropy rather than reduce it, because it narrows you into a smaller group of users.

Brave intentionally randomizes canvas, WebGL, and AudioContext output on a per-site, per-session basis as a countermeasure — the values are close enough to a real render that most sites can't detect the noise, but different enough between visits that they can't be used as a stable identifier. This is why your Brave fingerprint hash will change on re-test even though your actual hardware hasn't. A stable hash across repeated tests in a "hardened" browser is actually a sign the anti-fingerprinting feature isn't active or has been disabled.

Yes, this is a known cross-browser fingerprinting technique. Even though canvas and WebGL hashes differ slightly between browser engines, hardware-level signals — CPU core count, installed system fonts, screen resolution, GPU model exposed via WebGL renderer string, and timezone — are often identical across browsers on the same machine. Research groups have demonstrated cross-browser correlation with well over 90% accuracy using exactly this kind of hardware-signal overlap, independent of any cookie or login. Run our canvas fingerprint test in each browser to compare the individual hashes yourself.

In the EU, ePrivacy rules generally treat fingerprinting for tracking purposes the same as cookies — it requires consent under most regulator guidance, including from France's CNIL and the UK ICO, because the outcome is the same: identifying a device across visits without the user's clear opt-in. In practice, most cookie consent banners only mention cookies and say nothing about fingerprinting, which means a huge share of sites collecting these signals are doing so without the consent their own banner implies they're asking for.

A high score means your combination of signals is statistically rare enough to reliably re-identify you if a site chooses to fingerprint you — it doesn't confirm any specific site is doing so at this moment. Think of it as measuring exposure, not an active attack: a high score tells you how easy tracking would be if attempted, which is exactly the information you need to decide whether to act on the protection steps below.