User Agent Parser

Every browser introduces itself to every website with a "user agent" string — a dense line of text naming the browser, its version, the operating system and the device. This free user agent parser decodes it into plain English. Your own user agent is already filled in and parsed below, which answers the classic support question "what browser are you using, exactly?" — and you can paste any other UA string (from your analytics, server logs or a bug report) to identify what produced it: genuine Chrome on Windows, Safari on an iPhone, or a bot pretending to be both. Everything is parsed in your browser; nothing is sent anywhere.

How to use this tool

  1. See your own user agent parsed on arrival.
  2. Or paste any UA string from logs or analytics.
  3. Read browser, OS, device type and engine.
  4. Spot bot and crawler tokens flagged in the string.

Worked example

A log line claiming 'Chrome on Windows' parses in a second — and the crawler token hiding in it explains why that 'visitor' never bought anything.

Frequently asked questions

What is a user agent, in plain words?

It is the introduction line your browser sends with every web request: roughly "I am Chrome 126, running on Windows 10, on a desktop". Websites use it to serve the right layout and to count visitors by browser and device in analytics. The format is a historical pile-up — nearly every modern browser's string contains the words "Mozilla/5.0" and mentions of engines it does not use — which is exactly why a parser is handy.

Why does my Chrome string mention Safari and Mozilla?

Compatibility history. In the 1990s sites served their good pages only to browsers naming themselves "Mozilla"; rivals copied the name to get the good pages, then added their own tokens on top. Layers accumulated: Chrome's string mentions Mozilla, AppleWebKit (its engine ancestor), KHTML and Safari, before finally saying "Chrome". The parser untangles the pile-up by checking the tokens in the right order — real identity tokens like Edg, OPR and Chrome before the legacy mentions.

Can a user agent be faked?

Completely — it is just text the sender chooses. Bots routinely claim to be Chrome on Windows; privacy tools randomise it; browsers let developers override it in device-testing modes. Treat UA data as a helpful hint, never proof: analytics segments and log analysis are fine uses, security decisions are not. That is also why the parser flags known bot and crawler tokens when they are honestly present in the string.

Are user agent strings going away?

They are shrinking. Chrome and others are "freezing" parts of the string (the exact OS and browser build no longer change) and moving detail into a newer mechanism called User-Agent Client Hints, to reduce passive fingerprinting. The string will remain for years for compatibility, and what remains — browser family, platform class, mobile or not — is still exactly what support and analytics usually need. This parser reads the classic string.

Related tools: see the network side of your visit with the IP address lookup, and measure your connection with the internet speed test.

Related tools

Back to the tool directory