What device am I using?
No website is told what kind of device you have. It works the category out from clues, mainly whether you have a touchscreen and what your user agent claims, and the clues are noisy. A touchscreen laptop and a large tablet look almost identical from here, which is why responsive design leans on screen size instead.
A summary of the device your browser reports is shown above: whether it looks like a desktop, mobile or tablet, and the capability hints that back that up, such as touch support and platform. Browsers do not hand over a clean device name, so this is inferred from several signals together.
The inference blends the user agent's platform tokens, the mobile Client Hint where available, touch support, and screen size. No single one is decisive (a touchscreen laptop muddies the water), so treating device type as a confident guess rather than a fact is the honest approach.
This matters because sites branch heavily on it: mobile users get larger tap targets and simplified navigation, desktop users get denser layouts and hover interactions. Getting the guess wrong leads to the familiar frustration of a mobile layout on a big screen or vice versa.
How a website reads this
This is the exact call a page makes. Nothing else is involved, and nothing is sent anywhere: the value is read in your browser and displayed to you.
navigator.maxTouchPoints with navigator.userAgentThere is no API that says "phone" or "laptop". The category above is inferred from touch capability, the platform tokens in the user agent and the screen dimensions. It is a guess, and it gets touchscreen laptops and tablets in desktop mode wrong regularly.
- Needs your permission
- No, it is read silently
- Can it be faked
- Yes. Every input to the guess can be faked, and the browser's own "request desktop site" setting fakes it for you.
- Contribution to a fingerprint
- About 1 bits, our own estimate. Our estimate. Three broad categories cannot narrow much on their own; the inference is worth more as a cross-check against other signals than as a signal in itself. How we estimate this.
What you should know
Why sites collect it
Sites use device type to choose layouts, input assumptions (touch versus mouse), and features. App download pages use it to offer the right store link, and analytics use it to understand their audience mix.
Privacy implications
Device type is a coarse, low-entropy signal on its own, since it only sorts you into a few broad buckets. Its privacy relevance comes from the underlying signals (user agent, touch, screen) that combine into it, each of which is discussed on its own page here.
How to change or hide it
You influence device detection mainly through the user agent and touch settings. Requesting a desktop site on mobile flips the reported type, and privacy browsers normalise these signals. There is no single device-type toggle.
Frequently asked questions
How does a site know if I am on mobile?
It combines your user agent, a mobile hint, touch support and screen size into a best guess. No single signal is definitive, which is why touchscreen laptops can confuse it.
Can I make a site treat my phone as a desktop?
Yes. Most mobile browsers have a request-desktop-site option that changes the user agent, which usually flips the detected device type.
Sources
Browser behaviour changes and vendors cap these values differently, so the claims above that are not self-evident are checked against the specifications and vendor documentation below.
Last checked on . How these values are measured and estimated is set out on the methodology page.
Something wrong here? Tell us what your browser reports and it gets corrected.