About the User Agent Parser
User agent strings are notoriously cryptic, layered with historical compatibility tokens. This parser reads one and reports the browser and version, the rendering engine, the operating system and the device class (desktop, mobile, tablet or bot). Leave the input empty to parse your own browser's user agent.
The layering is deliberate and very old: almost every browser still opens with Mozilla/5.0, and every Chromium browser also carries AppleWebKit, KHTML, like Gecko and Safari tokens so that decades-old server-side sniffing keeps working. Order therefore decides the answer. The distinctive token is tested first, OPR/ and Edg/ ahead of Chrome/, and Chrome/ ahead of Safari/, or else Edge reports as Chrome and Chrome reports as Safari.
One result surprises people. Safari on an iPad has described itself as Macintosh since iPadOS 13, so a tablet can legitimately look like a desktop, and nothing in the string can recover the difference. Modern browsers also freeze the version numbers they publish here, which is why a reported version can lag the real one.
Reading a support ticket or a server log and reproducing the exact browser and platform a bug was filed against is what it is for. The detection covers the major browsers and platforms; niche browsers may be reported by their underlying engine, and crawler tokens are labelled as Bot, which pairs with the robots.txt Generator when you decide which of them to allow. Nothing is sent anywhere.
How to use
- Paste a user agent string, or leave the input empty for your own.
- Read the browser, OS and device.
Common questions
- How do I see my own user agent?
- Leave the input empty and press Parse.
- Why does Chrome appear in a Safari string?
- User agents carry compatibility tokens for historical reasons. The parser checks the distinctive tokens first to get the real browser.
- Does it detect bots?
- Common crawler tokens are recognised and reported as Bot.