What is a user agent?
A user agent is any piece of software that acts on your behalf on the web — usually a browser, but also search engine crawlers, feed readers, command-line tools like curl, and the HTTP libraries inside apps. Whenever a user agent requests a page, it sends a User-Agent header identifying itself: which product it is, which version, what rendering engine it uses, and what operating system it runs on. The string grew organically over thirty years, which is why a modern Chrome string still begins with Mozilla/5.0 and name-checks both KHTML and Safari — every browser once claimed compatibility with its rivals to avoid degraded pages, and the claims never went away.
What websites do with it
Servers and scripts use the user agent for content negotiation and compatibility: serving a mobile layout to phones, polyfilling features an older browser lacks, or warning that a browser is unsupported. Analytics platforms aggregate it to report browser and device share. Security systems look at it — alongside other signals — to spot obvious bots, scrapers, and credential-stuffing tools, whose user agents are often missing or generic. None of these uses require the string to be truthful, and it is not: it is self-declared text that any client can set to anything, which is why serious systems never trust it on its own.
Privacy and fingerprinting
On its own, a user agent identifies a browser family, not a person — millions of people share the same string. The privacy problem is combination. A user agent plus your screen size, installed fonts, time zone, language list, CPU core count, and dozens of other small signals (several of which are shown in the table above) can narrow a visitor down to a very small group, or even a unique device, without any cookies. That is browser fingerprinting, and the details on this page are exactly the raw material fingerprinting scripts assemble. A mainstream browser with default settings keeps you in a larger crowd.
User agent reduction and freezing
Because of that fingerprinting risk, browsers are actively shrinking what the string reveals. Chrome has been rolling out User-Agent Reduction: the string now reports a fixed, frozen version number and a generic platform token, and sites that genuinely need specifics must ask for them through the User-Agent Client Hints API (Sec-CH-UA headers and navigator.userAgentData), where the browser can limit or refuse high-entropy values. Safari has long frozen parts of its string, and Firefox caps and simplifies version details too. The industry direction is the same: less passive information broadcast to every server, more detail only on request.
How to change your user agent in developer tools
Developers legitimately need to test how a site behaves for other browsers, and every major browser can override its user agent temporarily. In Chrome and Edge, open Developer Tools, open the Command Menu (Ctrl+Shift+P or Cmd+Shift+P), run "Show Network conditions", and in that panel untick "Use browser default" under User agent and pick or paste a string. Firefox offers the same through Responsive Design Mode (Ctrl+Shift+M), which has a user agent field. An override usually lasts only for that tab or session, and sites may still detect the real browser through JavaScript feature tests.
How to use this page
Everything above the article happens automatically: the string your browser reports, a heuristic breakdown into browser, operating system, device type, and bot likelihood, the navigator properties any website can read, and a server-side echo for comparison. The Copy button is handy when a support team asks what browser you are using — paste them the whole string. Testing a user agent override? Reload this page and watch the parsed values change.
Frequently asked questions
Can a website tell exactly who I am from my user agent?
Not from the user agent alone — it names software, not people. But combined with your IP address, logged-in accounts, and fingerprinting signals, it contributes to identification.
Why does my Chrome string mention Mozilla and Safari?
Historical compatibility. In the 1990s and 2000s, sites served their best pages only to browsers whose strings matched ones they recognised, so every new browser claimed to be Mozilla, then added KHTML, Gecko, and Safari tokens to pass those checks. The tokens stuck, and parsers now read from the end of the string backwards to find the honest part.
Why did the parser get my browser or version wrong?
Because parsing is heuristic and vendors now deliberately freeze or reduce version details in the string. A reported version of, say, Chrome 120 may simply be the frozen placeholder your real, newer browser sends. Where accuracy matters, feature detection beats string parsing every time.
Is it legal or safe to change my user agent?
Changing your own browser's string for testing or privacy is fine and built into developer tools. Using a fake user agent to impersonate another service, evade a site's access controls, or scrape against its terms is a different matter — the string being editable does not make every use acceptable.
Explore more tools
- What Is My IP — see the public IP address you are connecting from.
- DNS Lookup — query A, AAAA, MX, TXT and other DNS records.
- SSL Checker — inspect a site's TLS certificate and expiry.
- Port Checker — test whether a TCP port is reachable.
- Speed Test — measure your connection's download and upload speed.
- Subnet Calculator — work out network ranges from a CIDR block.
- Base64 Encoder & Decoder — encode and decode Base64 with UTF-8 support.
- Timestamp Converter — convert Unix timestamps to dates and back.
- JSON Formatter — format, validate and minify JSON.
- Password Generator — create strong random passwords locally.
- All NetTools — browse the full tool directory.