How to parse a User-Agent string
Paste a User-Agent string into the box — or click Use my browser's UA to load your own — and the breakdown appears instantly below. The parser reports the browser and its version, the layout engine (Blink, Gecko, WebKit, Trident), the operating system and version, the device type (desktop, mobile or tablet) and whether the string looks like a bot or crawler. It is built for the everyday tasks of reading a log line, debugging device-specific behaviour, or checking what a client is reporting, without sending that data to a third-party API.
What a User-Agent string contains
A User-Agent is a single header sent with every HTTP request. By long-standing convention it begins with Mozilla/5.0 for almost every browser, followed by a platform section in parentheses (operating system, CPU architecture, device hints) and one or more product tokens such as Chrome/124.0.0.0 or Firefox/126.0. WebKit and Blink browsers also carry AppleWebKit/537.36 and a trailing Safari/537.36 token for compatibility. Because the format grew organically over decades, it is messy and full of legacy tokens — which is exactly why a careful parser is more reliable than eyeballing it.
Why matching order matters
The tokens overlap, so the order of checks is what makes parsing correct. Microsoft Edge includes Chrome and Safari in its string and adds Edg/; Opera adds OPR/ on top of the Chrome tokens; Chrome on iOS reports CriOS while still resembling Safari. This tool tests the most specific product first — Edge and Opera before Chrome, Chrome before Safari, the mobile iOS variants before the desktop ones — so a Chrome string is not misreported as Safari and an Edge string is not misreported as Chrome. Each field is matched independently, so an unusual combination still yields a sensible best-effort result rather than a wrong one.
Bot and crawler detection
A large share of web traffic is automated, and bots advertise themselves in the User-Agent. A separate rule set recognises the major search-engine crawlers (Googlebot, Bingbot, DuckDuckBot, YandexBot, Baiduspider), social-media link scrapers (such as facebookexternalhit and Twitterbot), and generic markers like bot, crawler and spider. When one matches, the result flags the request as automated and names the bot where possible, which is useful for filtering analytics or understanding unexpected log entries.
Why parse User-Agents locally
User-Agent strings sit in your server logs alongside IP addresses and request paths, so they are part of your operational data. Pasting log lines into an online parser or an AI assistant sends that data to someone else's server and invites the kind of confident-but-wrong answer language models give on fiddly string formats. This tool applies a fixed, inspectable set of regular expressions in your browser tab: the same input always yields the same structured output, instantly, offline, with nothing transmitted. Paste a whole batch of log lines and read them in private.
Frequently asked questions
- What is a User-Agent string?
- An HTTP header identifying the browser, version, layout engine and operating system of the client making a request.
- Why does Chrome's User-Agent mention Safari and Mozilla?
- Legacy compatibility — nearly all browsers include
Mozilla/5.0, and Blink/WebKit ones keep the Safari token. The parser checks the most specific product first. - Can it detect bots and crawlers?
- Yes — a dedicated rule set recognises common crawlers, scrapers and generic bot tokens and flags the request.
- Is the User-Agent sent anywhere?
- No. Parsing is local regular-expression matching, so nothing is uploaded.