Test your browser
See what a fingerprinter sees — and, more usefully, how many other people have the same result.
This is not your typical fingerprinting test pgae. This one tells you, accurately, what values you share with millions of other people and what values set you apart.
Each measurement below is re-implemented from the telemetry in Firefox and computes the value exactly the way Firefox did when we gathered our dataset. Your result is then looked up in the aggregate distributions published with the paper, covering 24,871,013 Firefox release users across Windows, macOS, Linux and Android. For each characteristic you get your value and what percentage of all measured browsers reported the same value.
Nothing leaves your browser. The measurements run locally and are compared against a static reference file that is part of this website. No values are uploaded, stored, or logged, and there is no server-side component to this page. The one exception is the Send diagnostics button at the foot of this page: if you press it and confirm, your diagnostic report is sent to us.
There are some limitations. Fingerprinters rely on side channels to determine some information about your computer — for example, what fonts you have installed. To more accurately measure that signal, we collected font data directly in C++, avoiding the side channel. As shown below, when we calculate a fingerprint simulating what a fingerprinter would see, the fingerprint we calculate is not based on the exact data they see, but on equivalent data — and, in the case of side channels, more accurate and reliable data. This, along with using a strict cryptographic hash function rather than fuzzy matching, provides us with an upper bound of their accuracy. It also means, however, that on this page we cannot calculate an equivalent fingerprint to the ones we simulated. So we cannot tell you accurately whether you are unique, or how many other users you share a fingerprint with. However, the most uncommon value you have below is definitively an upper bound of how common your fingerprint would be — and, as we show in the paper, it is most likely even rarer than that.
If something on this page looks wrong — a value that seems far rarer or commoner than it should, a measurement that fails, a result that makes no sense — please report it as an issue on GitHub. It helps to include your diagnostics: the Copy diagnostics button at the foot of the page puts them on your clipboard. GitHub issues are public and the diagnostics contain your measured values, so if you would rather not post them, email them to tom@ritter.vg instead.
flowchart TD subgraph collection[" "] priv["Privileged JS"] unpriv["Unprivileged JS"] cpp["C++"] end priv --> telemetry unpriv --> telemetry cpp --> telemetry telemetry["Telemetry"] --> simulation["Simulation"] simulation --> akamai["Akamai Fingerprint"] simulation --> signifyd["SignifyD Fingerprint"] simulation --> fpjs["FingerprintJS Pro Fingerprint"]How common your value is in the population it is compared against:
- over 40%
- 10–40%
- under 10% — more identifying
- not in the published data at all
Test tools
Re-running measures everything again without reloading, which is what you want after changing a browser setting. Clearing storage forgets the values this page remembered from earlier visits, so the Protected Attributes section starts over. Copying diagnostics puts what the page measured, and how, on your clipboard — it includes your measured values and nothing else about you, and it goes nowhere on its own.
This sends the report below to the authors of the paper, to help us find where this page gets things wrong. It contains everything the page measured about your browser, which together can identify it, and nothing else. It is stored privately, without your IP address, and used only to improve this page.