What is my user agent?

Your user agent and Client Hints right now, plus a parser for any UA string: browser, version, OS, device, engine and bots.

Your user agent

Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; [email protected])

ClaudeBot (Anthropic)

Unknown

Bot / script

Device

WebKit

Engine

This looks like a bot or a script. Keep in mind a User-Agent is easy to fake — a real Googlebot is verified by IP (reverse and forward DNS).

Client Hints

Chrome, Edge, Opera and other Chromium browsers send precise details — OS version, model, architecture — not in the User-Agent string but via Client Hints: short Sec-CH-UA headers and the navigator.userAgentData JavaScript API.

Values from JavaScript (getHighEntropyValues)

…

What is a user agent and how to find yours

A user agent is the string your browser sends to every website: its name and version, operating system and device type. This page shows your user agent as soon as it opens, parses it, and lists Client Hints — the precise details Chrome no longer puts in the string.

Your user agent on page load, one-click copy

Client Hints: real OS version, model, architecture

Browser, version, OS, device, engine

Detects search and AI crawlers and HTTP clients

How to Use

Open the page — your user agent and its breakdown appear right away
Check the Client Hints block if you use a Chromium browser
To parse another string (e.g. from server logs), paste it and click “Parse”

FAQ

Windows 11 and Windows 10 both send “Windows NT 10.0”, and Chrome’s User-Agent reduction freezes other details too. Only Client Hints can tell them apart: a platformVersion of 13 or higher means Windows 11.

They replace the long UA string in Chromium browsers. By default a site only gets the brand, major version, platform and a mobile flag (the Sec-CH-UA headers). Details such as OS version, model and architecture must be requested separately. Firefox and Safari do not support Client Hints.

Yes — it is just a header. Extensions, DevTools device mode, scripts and scrapers change it freely. That is why a user agent alone cannot prove a bot is genuine: a real Googlebot is verified by IP with reverse and forward DNS.

Since 2022 Chrome has been reducing the user agent to make tracking harder, so only the major version is left in the string. The full build number is available via Client Hints (fullVersionList).

Yes. Your browser sends it with every request, together with your IP address and languages. Sites use it to serve a mobile layout or suggest the right download, and analytics tools use it to count browsers and devices.

What a user agent string is made of

The User-Agent is a header your browser sends with every request. A typical Chrome on Windows string looks like this (the version number is just an example):

Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/140.0.0.0 Safari/537.36

Part	Meaning
`Mozilla/5.0`	A historical compatibility token that almost every browser sends
`(Windows NT 10.0; Win64; x64)`	Platform: operating system and architecture
`AppleWebKit/537.36 (KHTML, like Gecko)`	The engine and its lineage; Chrome froze this number long ago
`Chrome/140.0.0.0`	Browser and major version
`Safari/537.36`	Another compatibility token — it does not mean Safari

That is why a browser is identified by the combination of tokens in the right order, not by the first word: Edge adds Edg/, Opera adds OPR/, Chrome on iPhone says CriOS/. Every browser on iOS runs on WebKit, even when it is called Chrome or Firefox.

How to find your user agent

The easiest way is this page: your user agent appears as soon as it loads and you can copy it with one click. Other options:

run navigator.userAgent in the browser console (F12 → Console);
open DevTools → Network, pick any request and find the User-Agent header under Request Headers.

If the JavaScript value differs from what our server received, the tool flags it — usually an extension or DevTools device emulation is changing it.

User-Agent reduction in Chrome

A full UA string carries enough detail to help tell users apart, so since 2022 Chrome has been freezing parts of it step by step. Other Chromium browsers (Edge, Opera, Brave and more) do the same.

Detail	In the UA string	Where to get the real value
Browser version	Major only: `140.0.0.0`	Client Hints: `fullVersionList`
Windows 10 vs 11	Always `Windows NT 10.0`	`platformVersion`: 13 or higher means Windows 11
macOS version	Always `Intel Mac OS X 10_15_7`	`platformVersion`
Android phone model	`Android 10; K`	`model` and `platformVersion`

Safari freezes the macOS version in its string too, and Firefox does not support Client Hints at all, so those browsers will not tell a website the exact OS version.

Client Hints: how sites get the precise details

By default Chromium browsers send short headers: Sec-CH-UA (brands and major versions), Sec-CH-UA-Mobile and Sec-CH-UA-Platform. Anything else must be requested — with an Accept-CH response header or from JavaScript:

navigator.userAgentData.getHighEntropyValues(
  ['platformVersion', 'model', 'architecture', 'fullVersionList']
).then(console.log);

The Client Hints block on this page shows both the headers our server received and the result of that call.

How to change your user agent for testing

In Chrome: DevTools → ⋮ menu → More tools → Network conditions → untick “Use browser default” under User agent, then pick or paste a string. Device mode (the phone icon in DevTools) changes the UA as well. It is handy for checking a mobile layout or redirect, but a site can still recognise your real browser through Client Hints and other signals.

Bot user agents and how to verify them

Search and AI crawlers name themselves in the UA: Googlebot, bingbot, YandexBot, GPTBot, ClaudeBot, PerplexityBot. Paste a string from your server logs into “Parse another user agent” and the tool tells you whose bot it claims to be. Never trust the string alone, though: a real Googlebot is verified by IP — reverse DNS must point to googlebot.com or google.com, and a forward lookup of that name must return the same IP. Use reverse DNS for that, and the robots.txt checker if you want to allow or block crawlers.