User Agent Parser: Browser, OS, Device and Bot Check

What a user agent is

The user agent is a line of text your browser sends to every website with every request. It names the browser and its version, the rendering engine, the operating system and often the device. Sites use it to serve the right page, count visitors and spot automated traffic.

User agent strings are messy. Almost all of them start with "Mozilla/5.0" for historical reasons, and Chrome's string mentions Safari and Gecko. A parser untangles the pieces so you do not have to.

What this parser shows

Browser name and version.
Rendering engine, such as Blink, WebKit or Gecko.
Operating system and version.
Device type, vendor and model when the string includes them.
Whether the string looks like a known bot or crawler.
UA Client Hints from your own browser, when it supports them.

User agent versus Client Hints

Browsers based on Chromium are reducing the detail in the user agent string, for example by freezing the version of the operating system. In its place they offer UA Client Hints, a structured set of values that sites can request. This page reads them with navigator.userAgentData, so you can compare what the classic string says with what the hints say. Firefox and Safari do not provide Client Hints.

Limits and privacy

A user agent is just text and anyone can change it, so parsing it tells you what a client claims to be, not what it is. A scraper can pretend to be Chrome, and a developer tool can pretend to be a phone. Bot detection here only recognises well-known crawlers that announce themselves, such as Googlebot and Bingbot.

Parsing happens in your browser using a bundled open-source library under the MIT licence. The string you paste is not sent anywhere.

Example: what a parsed string looks like

A string such as "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36" parses to Chrome 120, the Blink engine, Windows 10 or 11 and a desktop device. The same Windows version token, NT 10.0, is used by both Windows 10 and 11, so no parser can tell them apart from the string alone. Client Hints can, which is one reason they exist.

An iPhone string names iOS and Safari, and a Googlebot string names the crawler and links to Google's page describing it. Paste any string into the box to see how it breaks down, whether it came from a log file, a bug report or a support ticket.

When to use it

Reading access logs to see which browsers and devices visit a site.
Reproducing a bug report that only includes a user agent.
Checking whether a request that claims to be a search engine crawler is formatted like one.
Seeing what your own browser reveals before and after changing privacy settings.

Frequently asked questions

Why does my Chrome say Safari and Gecko?

For compatibility. Early websites served better pages to certain browsers, so every new browser copied the others' names to get the good version. The habit stuck.

Can I change my user agent?

Yes. Browser developer tools and some extensions can override it. Changing it only changes what the site is told, not your IP address or other fingerprint details.

How do I tell if a visitor is a bot?

Good bots name themselves in the user agent and can be verified by reverse DNS. Bad bots copy a browser string, so the user agent alone cannot catch them.

Why is the device unknown?

Desktop browsers rarely reveal the device. Phones and tablets usually include a model, though newer ones often hide it.

Is the pasted string sent to a server?

No. The parsing library runs in your browser and the page makes no request with your input.

Parse a user agent string