Basics · As of November 18, 2026 · By

What a website can learn about your browser

As soon as you open a website, its server sees some details about your connection. The site can also ask your browser for more information. Some kinds of access, such as using your camera, require permission. This guide explains which data falls into each group. Sources are at the end of each section.

At a glance

  • When you visit a website, its server sees the incoming IP address, the time of the request and the details your browser sends with it.
  • Depending on the browser, a site can collect details such as screen size, language, time zone and graphics information without asking. Together, these details can form a browser fingerprint.
  • A website needs permission to access your precise device location, camera or microphone. It can read a file if you select it in the file picker, for example.
  • During a normal visit, your browser doesn't give the site general access to your browsing history, saved passwords, files or content from other websites.

Try it yourself: What can a website see about you?

This guide covers ordinary websites in desktop browsers, including on Windows. What a site actually receives depends on factors such as your browser version, operating system, settings and permissions you've already granted.

1. What your connection reveals

The IP Check shows which address reaches our server and what that address can tell you. The Blacklist Check shows whether the list we use contains an abuse report for the address.

Data sourceWhat's visible or can be inferred
IP address and timeThe server or a service in front of it, such as a content delivery network (CDN), sees the incoming IP address and the time of the request. IP databases can associate the address with a network operator, an autonomous system number (ASN), a country or an approximate region. They may also point to hosting, a proxy or a VPN. The match can be wrong. More than one person or device can use the same public IP address.
HTTP requestThe server sees the requested address, including the path and query parameters. It also receives whatever headers your browser sends. These can include User-Agent, Accept-Language, Client Hints, privacy signals, a relevant cookie and, depending on the request, Referer or Origin. Your browser doesn't send the part after the # in the HTTP request. Scripts on the page can still read it.
TLS, HTTP/2 and HTTP/3The server or service that accepts the connection can examine details of the TLS handshake and the protocols used. These can produce fingerprints such as JA3 and JA4. The value can change with the handshake, and different clients can have the same value. A QUIC connection ID belongs to the connection.
Additional network testsA DNS test can give you a clue about which service resolves domain names. WebRTC may reveal additional addresses or connection details, depending on the browser, network path and privacy settings. The Network Information API can estimate latency or download speed and report a connection class. For instance, effectiveType: "4g" can also appear over Wi-Fi.

2. What a page can measure automatically

As a page loads, it can use JavaScript or CSS to find out the size of your browser window, for example. It can show the measurement on the page or keep it in the browser. The measurement only goes to a server if the page or an embedded service sends it.

AreaExamples
Browser and operating systemA site can ask for the browser's brand, version and platform, and see whether the browser identifies itself as mobile. Some browsers provide more detailed Client Hints through getHighEntropyValues(), such as full browser or platform version, architecture, bitness or device model. Supported JavaScript and CSS features, PDF support and differences between browser engines provide further clues. In modern browsers, navigator.plugins usually contains standardized PDF entries or is empty.
Language and regionA site can read your preferred languages, time zone, UTC offset and daylight saving time changes. Depending on the API, it may also learn how your browser formats dates and numbers and which calendar or numbering system it uses. These values can reflect your browser or system settings.
Display and input devicesYour browser can report screen and window size, pixel density, color depth, color space, HDR support and the number of touch points, among other details. A site can also check which type of pointer is available and whether it supports hovering. CSS media queries can pick up preferences such as dark mode, forced colors, higher contrast or reduced motion.
Hardware and performanceA site can ask how many logical processors the browser reports. Some browsers also give a rough estimate of device memory. Test calculations can show how long the browser takes to do specific work. Navigation Timing and Resource Timing give timings for the page request and its resources. System load, power saving and browser protections can affect these values.
Graphics and renderingA site can ask the browser to draw sample images, text or emojis on a canvas. It can also measure SVGs and layouts or use WebGL and WebGPU to ask which graphics features, limits and adapter details the browser reports. The result depends on factors such as the browser, operating system, fonts, graphics software and hardware. If two hashes match, that test produced the same hash value.
Fonts, speech and audioThe width or shape of rendered characters can give a site clues about available fonts. Depending on the browser, it may also check which text-to-speech voices are available. In a common audio test, the browser generates a signal in memory and analyzes it. The test doesn't need a microphone.
Media, battery and storageA site can check which audio and video formats your browser supports and whether it's likely to play them smoothly and efficiently. Depending on the browser, the site may also be able to see battery level, charging status, and estimates of how much storage the site uses and how much is available to it.
Privacy and permission signalsDepending on the browser, a site can check signals such as Global Privacy Control and the status of individual permissions. A status of “denied” means access is blocked in that situation. The status alone rarely tells the site which setting caused it. The browser may reveal limited details about the type of camera or microphone before you grant access.

3. What a website can observe during your visit

Sources

4. What requires your permission or a selection

FeatureWhat the site can receive
Device locationOnce you allow location access, the site can receive coordinates, a timestamp and an estimate of accuracy. Depending on the device, it may also get altitude, direction of travel and speed. The position can come from GPS, Wi-Fi, mobile networks or other sources. Permission for a website is separate from the Windows location service setting. A rough location can still be estimated from your IP address independently of either setting.
Camera and microphoneThe site receives video or audio from the device you've allowed it to use. Depending on the access granted, it may also get the device name, a browser-assigned device ID, resolution or sample rate. A site can ask for location, camera or microphone access as soon as it loads. If an earlier permission still applies, the browser may skip the prompt.
Screen sharing and displaysThe site receives the image from the tab, window or screen you select. It may also receive audio if you explicitly share it. To start a new capture, you'll need to choose what to share again. Some browsers have a separate permission for detailed information about multiple displays.
Files, folders and clipboardA website can read the name, size and contents of files you select. That includes embedded data such as GPS coordinates in a photo. If you share a folder, the site may also be able to access files within that folder, depending on the browser. Direct clipboard access depends on the browser's rules for focus, user actions and permissions.
USB, Bluetooth, HID and serial devicesIf your browser supports the interface, the site can receive information and data from the device you select. What it can access depends on the device and interface. An existing permission may still apply. Granting access doesn't give the site a full list of connected or paired devices.
Local fonts and local servicesSome browsers can provide a list of installed fonts and their details after you grant permission. Newer versions of Chrome may ask for separate permission when a site tries to reach certain services on your computer or home network. The rules on which responses the site can read still apply. This permission doesn't give the site a list of your Wi-Fi networks.

Sources

5. How websites recognize or link visits

MethodWhat it can link
Cookies and site storageA site can retrieve an identifier it saved earlier. It can do this with cookies, localStorage, sessionStorage, IndexedDB or the cache, depending on the purpose. Web Storage is separated by origin, meaning the scheme, host and port. Cookies have their own domain and path rules. An HttpOnly cookie can reach the server even though JavaScript can't read it.
Third parties and linksAn embedded service can see that a page was loaded. Whether it can connect visits across sites depends on factors such as available identifiers, browser rules and the data it already has. Redirects and unique identifiers in links can also connect visits. A general parameter such as utm_source=newsletter usually identifies the campaign.
Technical fingerprintsAn operator can look at several features together, even if individual values change. A test tool's hash depends on which measurements it uses and how it processes them. A match can point to the same browser or computer, but it can also occur with different devices or people. Studies have even matched different browsers on the same computer under certain test conditions.
Linking a visit to an account or recipientA session cookie, an already linked browser identifier or a unique link can connect a page view to information the operator already has. That can happen as soon as the page loads. If you forward a personal link, the operator may attribute the click to the original recipient even if someone else clicked.

Sources

6. What a site can't normally access

During an ordinary visit, a site has no general access to your browsing history, password store, local files, Windows user data, content from other sites, or nearby Wi-Fi names and access points. Depending on the data, access is limited by separation between sites, operating system rules, browser permissions or what you choose to share. Browser extensions, browser sync and a compromised browser can have different access rules.

Get updates on new guides

Sign up for updates and we'll send you the “Sauber einloggen” checklist as a PDF.

Sign up