User-Agent Parser

Paste one user-agent string—or enable batch mode for one string per line—to parse browser name/version, operating system, rendering engine, device vendor/model/type, CPU architecture, and known bot signals. RFC 9110 defines User-Agent as a request field containing product identifiers and optional comments about the software originating a request. This tool reads those self-reported tokens; it does not contact the device or inspect the browser that submitted them.

• The bundled UAParser.js 1.0.41 ruleset applies its browser, engine, OS, device, and CPU regular-expression data in the browser.
• The result exposes versions plus device vendor and model when the pasted string actually contains enough information.
• A separate bot layer recognizes named tokens such as Googlebot, Bingbot, OAI-SearchBot, GPTBot, PerplexityBot, ClaudeBot, and Applebot, then applies a generic crawler keyword fallback.
• Use my User-Agent reads navigator.userAgent from this browser; batch mode parses one pasted string per line.

Treat every result as a clue, not verified identity. User-agent strings can be changed or spoofed, compatibility tokens can name several browsers, and reduced strings may omit versions or device detail. Unknown values remain Unknown, while an unrecognized non-mobile string falls back to Desktop. Bot detection is also heuristic: an unlisted or disguised crawler can be missed, and an ordinary product name containing a crawler keyword can be flagged. Client Hints and capability detection can provide different or more useful signals when you control the application.

Parsing happens in your browser as you type. The input is not sent to a parsing API, but user-agent values can contribute to fingerprinting when combined with other data. Avoid treating this output as authentication, authorization, fraud proof, or a substitute for capability detection.