What does the User Agent Parser pull out of a User-Agent string?

It reports the browser name and version, the rendering engine, the operating system and version, the device type (desktop, mobile, tablet, or bot), and the CPU architecture when the string includes one.

How does it tell Edge and Opera apart from Chrome?

Edge and Opera are built on Chromium, so their User-Agent still contains a Chrome token. The parser checks the Edg and OPR tokens before the generic Chrome token, so it labels them Edge or Opera correctly.

How is the device type decided from a User-Agent?

It looks for signals in the string: iPad or tablet tokens mean tablet, a Mobile token or iPhone means mobile, a known crawler token means bot, and anything else falls back to desktop.

Can it detect bots like Googlebot?

Yes. It matches named crawlers such as Googlebot, Bingbot, and DuckDuckBot, plus generic bot, crawler, and spider tokens, and marks the device type as bot when it finds one.

Does the User Agent Parser read my browser's real User-Agent?

Only when you click "Use my browser's", which copies your navigator.userAgent into the box so you can inspect it. Nothing is sent anywhere — the value is parsed locally.

Where does the User-Agent string I paste get processed?

Entirely in your browser. The string never leaves your device and is not uploaded to ArrayKit, so you can safely paste UAs from private logs.

User Agent Parser Online

Break any User-Agent string into browser, engine, OS, device type, and CPU in your browser. The string stays on your device.

The User Agent Parser runs entirely in your browser. The User-Agent strings you paste — and your own browser's User-Agent — never leave your device and are not uploaded to ArrayKit.

Open the URL Parser

About User Agent Parser

The User Agent Parser turns a raw User-Agent string into readable fields so you can see exactly what a client claims to be. Paste any User-Agent header, or click "Use my browser's" to load your own, and it breaks the string into browser name and version, rendering engine, operating system and version, device type (desktop, mobile, tablet, or bot), and the CPU architecture when the string advertises one. It recognises Chrome, Firefox, Safari, Edge, Opera, and Samsung Internet across Windows, macOS, Linux, Android, and iOS or iPadOS, and flags crawlers such as Googlebot. It is handy when you are reading server logs, debugging device-specific rendering, writing analytics rules, or testing feature detection. The whole string is parsed on your device.

Features

Extracts browser name and version from Chrome, Firefox, Safari, Edge, Opera, and Samsung Internet
Identifies the rendering engine (Blink, WebKit, Gecko, Trident, or Presto)
Reports the operating system and version across Windows, macOS, Linux, Android, and iOS or iPadOS
Classifies the device as desktop, mobile, tablet, or bot
Detects crawlers such as Googlebot, Bingbot, and other spiders
Surfaces the CPU architecture (x64, arm64, arm, or x86) when the string advertises it
One-click "Use my browser's" button loads your own navigator.userAgent
Copies the full parsed result as formatted JSON

How to use the User Agent Parser

Paste a User-Agent string, or click "Use my browser's" to load your own
Read the browser, engine, OS, device, and CPU fields in the table
Check the device badge to see if it is desktop, mobile, tablet, or a bot
Copy any single field, or copy the whole result as JSON

Example

Input

Mozilla/5.0 (iPhone; CPU iPhone OS 17_3 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.3 Mobile/15E148 Safari/604.1

Output

Browser: Safari 17.3
Engine:  WebKit
OS:      iOS 17.3
Device:  mobile

An iPhone Safari string resolves to Safari on iOS, classified as a mobile device.

Common errors & troubleshooting

Edge or Opera is reported as Chrome. — Chromium browsers keep the Chrome token for compatibility. This parser checks the Edg and OPR tokens first, so paste the complete string — a truncated UA that drops those tokens will read as Chrome.
An iPad shows up as a desktop instead of a tablet. — Modern iPadOS Safari sends a Macintosh User-Agent by default. Enable the site's mobile request mode or use the iPad's Request Mobile Website option to get an iPad string.
The browser or version field says Unknown. — The string may be from an in-app WebView, a niche browser, or a spoofed header with no recognisable token. The OS and device fields are still filled in when possible.
A real visitor is flagged as a bot. — Some libraries and preview fetchers include bot-like tokens (crawler, spider, HttpClient). Confirm against the raw string; the parser only reports what the User-Agent advertises.

Frequently asked questions

What does the User Agent Parser pull out of a User-Agent string?: It reports the browser name and version, the rendering engine, the operating system and version, the device type (desktop, mobile, tablet, or bot), and the CPU architecture when the string includes one.
How does it tell Edge and Opera apart from Chrome?: Edge and Opera are built on Chromium, so their User-Agent still contains a Chrome token. The parser checks the Edg and OPR tokens before the generic Chrome token, so it labels them Edge or Opera correctly.
How is the device type decided from a User-Agent?: It looks for signals in the string: iPad or tablet tokens mean tablet, a Mobile token or iPhone means mobile, a known crawler token means bot, and anything else falls back to desktop.
Can it detect bots like Googlebot?: Yes. It matches named crawlers such as Googlebot, Bingbot, and DuckDuckBot, plus generic bot, crawler, and spider tokens, and marks the device type as bot when it finds one.
Does the User Agent Parser read my browser's real User-Agent?: Only when you click "Use my browser's", which copies your navigator.userAgent into the box so you can inspect it. Nothing is sent anywhere — the value is parsed locally.
Where does the User-Agent string I paste get processed?: Entirely in your browser. The string never leaves your device and is not uploaded to ArrayKit, so you can safely paste UAs from private logs.

Related tools

URL Parser — Break a URL into its parts and list query parameters.
IPv6 ↔ IPv4 Converter — Convert between IPv4 and IPv6 (mapped/expanded/compressed) and extract embedded IPv4.
URL Builder & UTM Generator — Construct and validate fully encoded production URLs with standard analytics query flags (UTM) or custom query arrays.
CIDR / Subnet Calculator — Calculate network, broadcast, host range and mask from a CIDR block.
HTTP Status & MIME Types — Searchable reference of HTTP status codes and MIME content types.
Hash Generator — SHA-256 / SHA-1 / SHA-384 / SHA-512 via the Web Crypto API.