Paste a User-Agent string and see what it claims: the browser, OS and device (identified with ua-parser's community rules), a token-by-token account of why the string looks the way it does — every browser still claims to be Mozilla — and, just as important, which parts are frozen placeholders that stopped describing the real system years ago. Nothing you paste leaves your browser.
A User-Agent string is a chain of compatibility claims, each layer added so old sniffing code would serve the good version of a page. Mosaic begat Netscape (code name Mozilla), so Netscape said Mozilla/1.0; IE claimed Mozilla/4.0 (compatible; MSIE …) to get Netscape's pages; WebKit (a fork of KDE's KHTML) declared itself (KHTML, like Gecko) to get Firefox's; Chrome claimed Safari; and everyone still opens with Mozilla/5.0. Since 2021 the process has run in reverse: browsers now freeze or reduce what's left (exact versions, OS builds, device models) to shrink the fingerprinting surface, so several tokens are deliberate placeholders. This tool shows both layers: what the string says, and where it stopped being true.
Identification uses the ua-parser project's uap-core rules (Apache-2.0), compiled into this page — 1,270 patterns maintained by the community. This implementation passes all 18,000+ uap-core test fixtures and is differential-tested against the reference Python implementation; see how these tools are tested. Remember the limits: a User-Agent is an unverified claim. Anything — curl, a bot, a script — can send any string.