← All articles

Antidetect · 8 min read · 7/22/2026

Browser Fingerprinting Explained: What Websites Can Detect

Learn how browser fingerprints are assembled, tested, and used—and why changing your IP address alone does not prevent recognition.

Browser Fingerprinting Explained: What Websites Can Detect

Browser fingerprinting lets websites recognize a browser by combining details about its software, hardware, and configuration. Unlike cookies, the resulting identifier does not need to be stored on your device. This makes fingerprinting useful for fraud prevention but difficult to control when it is used for cross-site tracking.

This guide explains which signals matter, how fingerprinting works, what changing an IP address can and cannot do, and how privacy browsers, VPNs, proxies, and antidetect browsers differ.

What is browser fingerprinting?

Browser fingerprinting is a technique for collecting observable characteristics from a browser and device, then combining them into a profile. One characteristic—such as screen resolution—is rarely unique. A large set of characteristics can be distinctive enough to recognize a returning browser or place it within a small group.

Websites can collect some signals through ordinary HTTP requests. Others are available through JavaScript and browser APIs. A fingerprinting script may process these values locally, send them to a server, or convert them into a hash for comparison.

Fingerprinting can be:

  • Passive: Uses information already present in network requests, including the IP address, TLS characteristics, and HTTP headers.
  • Active: Runs code to query browser APIs, render graphics, inspect capabilities, or measure behavior.
  • Stateful: Works alongside cookies, local storage, or account logins.
  • Stateless: Attempts recognition without placing a persistent identifier on the device.

The technique is probabilistic rather than infallible. Software updates, hardware changes, browser settings, and network conditions can alter a fingerprint.

Which data points can websites collect?

A fingerprint may include dozens or hundreds of signals. Common categories include:

  • Browser and operating system: User-Agent data, platform, browser version, feature support, and client hints.
  • Display: Screen dimensions, available viewport, pixel ratio, color depth, and orientation.
  • Language and time: Preferred languages, locale, time zone, date formatting, and clock behavior.
  • Hardware: Reported processor threads, device memory, touch support, and graphics capabilities.
  • Fonts and media: Available fonts, supported codecs, audio formats, and media-device information where permitted.
  • Graphics rendering: Canvas and WebGL outputs, GPU vendor, renderer details, and supported extensions.
  • Audio processing: Small differences produced when the Web Audio API processes a generated signal.
  • Browser configuration: Cookie support, tracking preferences, permissions, installed extensions when inferable, and enabled APIs.
  • Network data: IP address, approximate location, autonomous system, DNS behavior, TLS handshake traits, and connection properties.
  • Behavior: Mouse movement, typing cadence, scrolling, navigation patterns, and timing measurements.

Access is not unlimited. Modern browsers restrict sensitive APIs, require permission for some data, and reduce or standardize certain values. Scripts may still infer information indirectly by checking whether features exist or observing how the browser renders content.

How a fingerprint is created and matched

The process usually follows four stages:

  • Collection: A script or server records available browser, device, rendering, and network signals.
  • Normalization: Values are standardized so minor formatting differences do not create unnecessary mismatches.
  • Scoring: Signals are weighted according to their stability and distinctiveness. A rare GPU-and-browser combination may contribute more than a common language setting.
  • Matching: The current profile is compared with previously observed profiles. The system returns an exact match, a confidence score, or a likely group.

Sophisticated systems do not depend on every value staying identical. They can tolerate expected changes, such as a browser version update, while considering whether the remaining attributes still point to the same device.

Consistency also matters. A profile may appear suspicious when its values conflict—for example, an IP address in one country paired with an implausible time zone, locale, and system configuration. Anti-fraud tools look for these mismatches as well as uniqueness.

Fingerprinting, cookies, and IP tracking compared

These methods overlap, but they are not interchangeable.

| Method | What identifies you | Stored on device? | Changes when IP changes? | Main limitation |

|---|---|---:|---:|---|

| Cookies | Random site-issued identifier | Yes | Usually no | Users can delete or block them |

| IP tracking | Public network address | No | Yes | Addresses may be shared or rotated |

| Browser fingerprinting | Combined browser and device signals | Not necessarily | Partly | Profiles can change and matches are probabilistic |

| Account tracking | Login or account identifier | Sometimes | No | Requires a known account or linked activity |

A VPN or proxy replaces the public IP address visible to a destination. It does not automatically change canvas output, screen size, browser APIs, fonts, language, or hardware characteristics. Consequently, an old and new IP may still be associated through the same fingerprint, cookies, or login.

The reverse is also true: changing browser characteristics while keeping accounts, cookies, or a stable IP can preserve other links.

Why websites use browser fingerprints

Fingerprinting is not inherently malicious. Its purpose depends on who deploys it and how the data is handled.

Common uses include:

  • Detecting account takeovers and payment fraud
  • Identifying automated traffic, scraping, or credential stuffing
  • Limiting repeated sign-ups, votes, trials, or promotions
  • Applying risk-based authentication
  • Measuring advertising reach and attributing conversions
  • Recognizing returning visitors when cookies are unavailable
  • Enforcing licensing, geographic, or platform restrictions

The privacy concern is that fingerprinting can be hard to observe, understand, or opt out of. A profile may also be shared with third parties or combined with IP data, browsing history, and account records. Consent and disclosure requirements vary by jurisdiction, so organizations should obtain legal guidance rather than assume fingerprinting is exempt from privacy rules.

How to test whether your browser is distinctive

Public fingerprint-testing pages can show exposed attributes and estimate whether a browser looks common or unusual. Treat the result as a snapshot, not definitive proof of tracking.

Use this checklist when evaluating a test:

  • Run it in a normal browsing session and record the visible signals.
  • Repeat it after restarting the browser to check stability.
  • Compare results with and without private browsing mode.
  • Test after enabling privacy protections or changing networks.
  • Check for contradictions among IP location, time zone, and language.
  • Remember that each test uses its own dataset and collection methods.

A claim that your fingerprint is “unique” only means unique within that service's observed sample at that moment. A different website may collect fewer signals, use different weights, or have a much larger comparison set.

Ways to reduce fingerprinting

No single setting guarantees anonymity. Effective defenses reduce exposed information, make users look similar to a larger group, or isolate identities from one another.

Prefer standardization over random changes

Privacy-focused browsers may standardize values, partition storage, limit APIs, or add carefully designed noise. Randomly changing individual settings can backfire: an unusual combination may become more distinctive or internally inconsistent.

Limit scripts and site permissions

Blocking unnecessary JavaScript can prevent many active techniques, although it can break websites. Review camera, microphone, location, notification, and other permissions. Remove extensions you do not need, since uncommon extension-driven behavior may add entropy.

Separate browsing contexts

Use separate browser profiles or containers for activities that should not be linked. Do not sign into the same accounts across identities, and avoid moving cookies or local-storage data between profiles.

Understand each privacy tool's role

  • VPNs: Encrypt traffic to the VPN server and replace the public IP, but do not inherently alter the browser fingerprint.
  • Proxies: Route selected traffic through another IP. They also do not automatically align browser characteristics.
  • Privacy browsers: Aim to block, partition, or standardize tracking surfaces for everyday privacy.
  • Antidetect browsers: Create multiple browser profiles with controlled fingerprint parameters, typically for authorized account management, QA, advertising operations, or research.

Antidetect tools are not invisibility switches. Poorly configured profiles can leak inconsistent values, retain identifying storage, or use an IP that conflicts with the profile. Use them only where permitted by the website, contract, and applicable law.

FAQ

Can private browsing stop browser fingerprinting?

Not by itself. Private mode mainly limits local history and clears session storage after windows close. It may change or restrict some signals, but websites can still observe many browser, device, and network attributes during the session.

Does changing my IP address change my fingerprint?

It changes one important network signal, not the complete fingerprint. Screen properties, graphics rendering, browser configuration, time zone, fonts, and other attributes may remain stable. Cookies and account logins can also connect sessions across IP changes.

Is disabling JavaScript enough?

Disabling JavaScript blocks many active collection methods, including common canvas and WebGL scripts. Servers can still observe IP, TLS, and HTTP characteristics. Many websites also require JavaScript, making complete disabling impractical for routine browsing.

Bottom line

Browser fingerprinting combines many weak signals into a stronger recognition profile. Cookies, IP addresses, logins, and behavior can reinforce that profile, so replacing an IP alone is not a complete defense. For everyday privacy, use a reputable privacy-focused browser, minimize extensions and permissions, and separate sensitive contexts. For legitimate multi-profile work, choose an antidetect browser that emphasizes coherent profiles, isolation, clear documentation, and compliance rather than promises of guaranteed anonymity.

Benchmark data

Figures below come from our own provider tests — the same dataset behind our provider reviews.

Request success rate

Successful responses across 12 target sites (higher is better).

Bright Data99.2%
Oxylabs98.7%
Decodo98.1%
SOAX97.3%
Webshare96.4%
Rayobyte95.8%
Average response time

Median time to first byte in seconds (lower is better).

Rayobyte0.5s
Webshare0.6s
Bright Data0.7s
Oxylabs0.8s
Decodo0.9s
SOAX1.1s
Proxy type coverage

Share of tested providers offering each network type.

  • Residential29%
  • ISP29%
  • Datacenter24%
  • Mobile19%