How to reduce what websites can see in your browser

Websites do not need your name to recognize you. They combine small technical details — like how your screen draws text, which fonts you have, and what language your browser reports — into a profile. This hub maps each type of signal to practical protections you can turn on today, and shows you how to verify they work.

Shield with radar scan representing browser privacy protection layers

How browser tracking actually works

When you open a page, your browser automatically sends a User-Agent string and other HTTP headers (language, privacy preferences, and in Chromium browsers, structured Client Hints). The page may also run JavaScript that reads your screen size, timezone, CPU threads, and runs fingerprint probes — hidden drawing tests (canvas), graphics checks (WebGL), audio processing tests, and font measurements.

Your IP address reveals approximate location and network provider. None of these signals alone is always unique, but combined they often form a stable identifier — even without cookies. Our analysis tool checks the same categories so you can see your real exposure and test changes before and after you adjust settings.

Diagram of browser fingerprint signals combining canvas, WebGL, and headers

Protection layers at a glance

Layer What it limits Deep-dive guide
Anti-fingerprinting Canvas, WebGL, audio, and font probes Anti-fingerprinting guide
Opt-out signals Global Privacy Control and Do Not Track headers GPC & DNT guide
Client Hints Structured browser data beyond the UA string Client Hints guide
Cookies & site data Stored identifiers and cross-site tracking Cookies guide
Locale & timezone Language headers and timezone fingerprinting Locale & timezone guide
Network anonymity IP-based location (not JS fingerprint) VPN guide
Maximum anonymity All probes plus network routing Tor Browser guide

Quick wins (start here)

Signal → protection map (simplified)

Deep-dive guides

Understanding your privacy score

The privacy score (0–100, higher is better) weights every inference card from your analysis — not just canvas or WebGL. Exposure signals add points; protective signals (blocked probes, GPC, canvas randomization) subtract. The Overview also shows estimated fingerprint entropy (~bits), core probe coverage (how many of the four baseline probes succeeded), and — when enough visitor data exists — how your probe richness compares to recent analyses on this site.

Cross-signal mismatches (for example VPN timezone vs browser timezone) can improve your score because inconsistent signals are harder for trackers to trust. See anti-fingerprinting for the full inference list.

Verify your protections

After changing settings, confirm they work — not just that you toggled a switch.

Run a free analysis to see which signals your browser exposes right now.

Check my browser now

Frequently asked questions

Do I need every protection listed here?

No. Start with quick wins: GPC, third-party cookie blocking, and anti-fingerprinting in your main browser. Add locale, Client Hints, or VPN adjustments based on what your analysis report flags. Tor is for high-stakes anonymity, not everyday browsing.

Will these settings break websites?

Stricter settings can cause CAPTCHAs, login issues, or broken layouts on some sites. Most guides explain trade-offs per browser. You can often use per-site exceptions instead of turning protections off globally.

What does probe richness mean?

Probe richness counts how many fingerprint probe classes succeeded during your visit (core canvas, WebGL, audio, and fonts plus extended math, media queries, speech voices, and secondary canvas draws — up to eight). When the benchmark has enough data, the Overview compares your richness to recent visitors. Fewer successful probes usually means less material for trackers to combine.

← Run the full analysis