User Agent Parser
Turn cryptic user-agent strings into readable facts: browser and version, rendering engine, operating system, device type, vendor and model, and whether the visitor is a crawler such as Googlebot, GPTBot or an SEO tool. Paste a single string, a list, or raw lines from an Apache or Nginx access log, get a summary of browsers, systems and bots, and download the results as CSV.
- Runs in your browser
- No sign-up
- Free to use
One per line, up to 5,000 lines. In access-log lines the last quoted field is used as the user agent. Your own browser’s string is filled in to start.
| Result | Browser | Engine | Operating system | Device |
|---|
How to use User Agent Parser
- Your own browser’s string is filled in; replace it or paste more.
- For server logs, paste the log lines directly.
- Read the parsed table and the summary.
- Download the CSV for further analysis.
User Agent Parser features
Browsers and apps
Chrome, Edge, Safari, Firefox, Samsung Internet, Opera, in-app browsers and WebViews.
Systems and devices
Windows, macOS, iOS, iPadOS, Android, ChromeOS, Linux, TVs and consoles, with phone models.
80+ crawlers
Search engines, AI crawlers, link-preview bots, SEO tools, monitors and HTTP libraries.
Log-file aware
Takes the user agent from the last quoted field of access-log lines.
Honest about reduction
Explains frozen Windows, macOS and Android values in modern strings.
Private and bulk
Up to 5,000 lines in your browser, with CSV export.
When to use User Agent Parser
- Finding out which browsers and devices visit a site, from raw server logs.
- Checking which bots, including AI crawlers, are fetching your pages.
- Debugging a bug report that includes a user-agent string.
- Learning what your own browser reveals about you.
User Agent Parser FAQ
What is a user-agent string?
A line of text that browsers and other HTTP clients send with every request to identify themselves. It usually names the browser, its engine, the operating system and sometimes the device model.
Why does every browser say “Mozilla”?
For historical compatibility. Early sites served better pages to Netscape (“Mozilla”), so other browsers copied the token, and later browsers copied each other’s tokens. The real browser is identified by the tokens further along.
Why does Windows 11 show as Windows 10?
Browsers froze the Windows version in the string at “Windows NT 10.0”. The exact version is only available through client hints, which the Browser Information Checker reads.
Can I trust the bot detection?
It recognises how bots identify themselves, but anyone can send any string. To confirm a real Googlebot or Bingbot, check that the visiting IP address resolves to the search engine’s domain and back, for example with the Reverse DNS Lookup.
Is anything uploaded?
No. All parsing runs in your browser, so log files never leave your computer.
Why is the device model missing?
Many browsers now send a reduced string without the model, such as “Android 10; K”, to make fingerprinting harder. The model is only available to sites through client hints.
Reading user-agent strings
A user-agent string is a self-description. A typical one from Chrome on Windows reads “Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36”: it claims Mozilla, WebKit, KHTML, Gecko, Chrome and Safari at once. Decades of sites checking for particular words led every browser to include the words others expected, so the useful information is in the order and combination of tokens rather than in any single one.
This parser works the way experienced developers read the strings. Bots and tools are recognised first, because they often imitate browsers. Then browsers built on Chrome or Safari, such as Edge, Opera, Samsung Internet and in-app browsers, are checked before Chrome and Safari themselves. The engine follows from the browser: Blink for Chrome-based browsers, Gecko for Firefox, and WebKit for Safari and every browser on iPhone and iPad.
Modern strings carry less than they used to. Chrome and its relatives have reduced theirs: minor version numbers are zeros, Windows always says 10.0, macOS is frozen at 10.15.7, and Android phones report “Android 10; K” instead of the real version and model. The parser shows these as uncertain rather than presenting a frozen value as fact. Sites that need the details can request them through User-Agent Client Hints.
Bots are a growing share of traffic. Search engines, AI training crawlers, link-preview fetchers used by chat apps, SEO tools, uptime monitors and simple HTTP libraries all identify themselves differently. Pasting a day of access-log lines gives a quick picture of who is fetching your pages; the summary counts bots separately from browsers so they do not distort the browser statistics.
Because user agents are self-reported, they should never decide security questions. Scrapers often pretend to be Chrome, and a request claiming to be Googlebot may come from anywhere. Use the string for statistics, compatibility workarounds and debugging, and use IP verification, authentication and rate limiting when it matters who is really on the other side.