Browser Fingerprint Test

What each signal is and who reads it

Any script on a page can read these values without a permission prompt, which is why fingerprinting keeps working after cookies are cleared or blocked. The W3C's note for specification authors, Mitigating Browser Fingerprinting in Web Specifications, explains how values like these combine to recognize a browser across sites.

• Canvas. The page draws text and shapes on an invisible canvas and hashes the pixels. Graphics hardware, drivers, font rendering and anti-aliasing all shift the output slightly.
• WebGL. The WEBGL_debug_renderer_info extension reveals the graphics vendor and renderer, and limits such as the maximum texture size narrow the hardware down further. Some browsers return a generic renderer name instead.
• Audio. An OfflineAudioContext renders a short tone through a compressor without playing anything. The floating-point result varies with the audio stack and processor.
• Fonts. Measuring a test string in a list of font families shows which ones are installed. Safari limits this to fonts that ship with the system.
• Screen and window. Resolution, available area, colour depth and pixel ratio.
• Locale. Time zone, UTC offset, languages and formatting locale. The IP and timezone check compares them with your IP location.
• Hardware. Logical processor count, rounded device memory in Chromium browsers, and touch points.
• Client hints. Chromium browsers expose brands, platform and a mobile flag by default and more detail on request through getHighEntropyValues(), part of the User-Agent Client Hints API.
• Automation flag. navigator.webdriver is true while a browser is controlled through WebDriver, as the W3C WebDriver standard requires.

The readers are bot-management and fraud scripts on login, sign-up and checkout pages, ad verification and analytics vendors, and each platform's own risk systems. For a longer tour of the sources, see 47 data points your browser exposes and the fingerprint detection guide.

Consistency matters more than rarity

A fingerprint that few other browsers share makes a browser easy to recognize again, which is what matters for tracking. For risk scoring, contradictions matter more. A rare but coherent browser looks like a person with unusual hardware. A common-looking browser whose values disagree looks assembled.

The checklist above looks for the contradictions that come up most:

• a Windows User-Agent with client hints that say macOS, or the reverse;
• a phone User-Agent on a device that reports no touch points;
• a software renderer such as SwiftShader or llvmpipe, which real desktops rarely use and virtual machines and headless browsers often do;
• Date and Intl reporting different offsets, which happens when only one of them was overridden;
• navigator.webdriver set to true, or a HeadlessChrome token in the User-Agent.

The 2024 FP-Inconsistent study measured this directly. Bots that altered fingerprint attributes to get past two anti-bot services often produced combinations that real devices do not, and rules written over those combinations cut their evasion rate by close to half (Venugopalan et al.). The same logic applies across the network boundary: a browser that says it is in Berlin should not exit from an address in Texas. The proxy detection test checks both sides at once.

How antidetect profiles change these

An antidetect browser keeps a separate set of these values for each profile, with its own cookies and storage, so several accounts do not share one fingerprint. Profiles take two broad approaches. Some substitute values taken from real device configurations, so canvas, WebGL and fonts fit together. Others add small random noise to canvas, WebGL or audio output. Noise breaks matching across sites, but it can be noticed, because the same drawing returns different pixels. The canvas reads row in the result shows whether that happens in your browser.

To check a profile:

1. Start it with its proxy assigned, open this page inside it and read the signals.
2. Confirm the checklist passes, then run the IP and timezone check in the same profile.
3. Open a second profile and compare. Hashes should differ between the two profiles, and each profile should give the same hashes every time you read it.

The antidetect browser reviews, the antidetect browser guide, mobile proxy setup for antidetect browsers and the best antidetect browsers list cover the tools themselves. The address behind each profile counts as much as the fingerprint: dedicated mobile proxies give each profile a real carrier line with its own SIM, used by one customer only.

Why there is no uniqueness score

A uniqueness figure needs a large set of other people's fingerprints to compare against. Projects that publish one, such as the EFF's Cover Your Tracks and AmIUnique, build that set from the fingerprints of their visitors. This page collects nothing, so a percentage here would be made up. Use those projects for a comparison with other people, and this page to see the raw values and whether they agree.