What does AmIUnique measure?

AmIUnique measures how identifying your browser's fingerprint is within its visitor population. The tool reads approximately 22 browser attributes and compares them against a dataset of millions of prior visitors maintained by INRIA, France's national institute for research in computer science. The output is a uniqueness percentage.

Is the AmIUnique uniqueness number accurate?

It is accurate within the AmIUnique dataset. The dataset skews toward European privacy-conscious visitors because of AmIUnique's referral base, so a '100% unique' result on AmIUnique may not mean the same browser is 100% unique on a general-population ad-tech tracker. The directional reading — 'your browser is highly identifying' or 'your browser is common' — is reliable.

What does Snitchtest add beyond AmIUnique?

Snitchtest reads a similar attribute set but presents per-attribute entropy contributions and specific defensive recommendations. AmIUnique shows you are unique; Snitchtest shows which attributes are leaking the most entropy and which browser settings would close those leaks. The two are complementary rather than redundant.

Is AmIUnique still maintained?

Yes. AmIUnique is maintained by the Spirals research team at INRIA. Its corpus is the research-grade reference dataset for fingerprint entropy studies in the academic literature, and its attribute measurements are kept current with modern browser APIs.

If you are trying to understand how identifying your browser is, yes. Run AmIUnique for the research-grade uniqueness score, then run Snitchtest for the per-attribute breakdown and the actionable defense recommendations. The two together give you both the reading and the prescription.

Snitchtest vs AmIUnique

AmIUnique is a research tool from INRIA's Spirals team — the canonical academic reference for browser fingerprint uniqueness measurements. Snitchtest is a per-user dossier that reads a similar attribute set and turns it into actionable privacy recommendations. Here is how the two differ, and why running both is the right move.

Updated 2026-04-22 · 5 min read · Snitchtest editorial

Summary comparison

Dimension	AmIUnique	Snitchtest
Primary output	Uniqueness % vs corpus	Per-attribute entropy + defenses
Data corpus	~3.5M visitors (research)	Live per-session read
Attribute count	~22 core attributes	~20 attributes + contextual
Canvas fingerprint?	Hash shown	Hash + entropy + defense
Audio fingerprint?	Yes	Yes
WebRTC leak check?	Limited	Full leak dossier
Defensive recommendations?	Minimal	Per-attribute guidance
Research provenance	INRIA / Laperdrix et al.	Independent / encyclopedia-register
Use case	"How unique am I?"	"What is leaking and what do I fix?"

What AmIUnique does well

AmIUnique is the single most-cited browser fingerprinting tool in academic literature. Pierre Laperdrix's 2016 paper introducing the corpus has been referenced in hundreds of subsequent fingerprinting studies, and the tool's dataset is the reference point for any measurement of how uniqueness evolves over time. When a researcher says "X% of Chrome users have a globally unique fingerprint," the underlying number almost certainly traces back to AmIUnique's or Panopticlick's corpus.

The tool's methodology is clean. It reads approximately 22 attributes — user agent, screen dimensions, installed fonts, WebGL renderer string, canvas hash, audio hash, timezone, platform, language, plugin list (legacy), and several others — and computes a per-visitor uniqueness percentage against the growing research corpus. The output is a single number you can understand at a glance: "your fingerprint is unique among X million visitors."

Where Snitchtest adds different value

AmIUnique tells you that you are unique. It does not tell you why, or what to do about it. A typical AmIUnique result says "your browser is unique among 3,500,000 visitors." Useful, but not actionable — which attribute is the most identifying? Which can you change? Which are effectively locked by your OS?

Snitchtest is built to answer those questions. The dossier shows each attribute, the estimated entropy it contributes (in bits), and specific defenses where they exist. Your canvas hash contributes roughly 9 bits? Here is the Firefox preference that neutralizes it. Your WebGL renderer string exposes your GPU model? Here is how Brave's per-session randomization changes that. The per-attribute breakdown is the part that closes the gap between "you are unique" and "here is what you do about it."

Which should you use?

"How unique is my browser?" — AmIUnique. Canonical research corpus, cleanest uniqueness percentage.
"Which attributes are leaking the most entropy?" — Snitchtest. Per-attribute entropy estimates.
"What specifically should I change to reduce uniqueness?" — Snitchtest. Defensive recommendations per attribute.
"I want to cite a fingerprint statistic in an article" — AmIUnique. Academic citation weight.
"I am building or evaluating a privacy browser" — Both. AmIUnique for the comparative benchmark against research corpus, Snitchtest for the attribute-level debugging view.

Credit where due

The field of browser fingerprinting measurement effectively exists because of Panopticlick (EFF, 2010) and AmIUnique (INRIA, 2016). Every subsequent tool — Cover Your Tracks, CreepJS, BrowserLeaks, Snitchtest — extends methodology that originated in that research. Snitchtest's per-attribute entropy estimation is a direct application of the approach Laperdrix's papers established.

Frequently asked questions

What does AmIUnique measure?: How identifying your browser's fingerprint is within its research corpus of millions of visitors, maintained by INRIA.
Is the AmIUnique uniqueness number accurate?: Within its corpus, yes. The corpus skews toward privacy-conscious European visitors, so directionally reliable but not population-general.
What does Snitchtest add beyond AmIUnique?: Per-attribute entropy contributions and specific defensive recommendations, in addition to the measurement.
Is AmIUnique still maintained?: Yes, by the Spirals research team at INRIA.
Should I use both?: Yes. AmIUnique for the research-grade score, Snitchtest for the attribute-level breakdown and defenses.