Bot detection test

What this test looks for

Automated browsers leave traces. A browser driven by WebDriver says so in navigator.webdriver, headless Chrome names itself in the User-Agent or opens without a window, and frameworks like Playwright, Selenium (ChromeDriver), PhantomJS and Nightmare leave their own objects on the page.

Bots also give themselves away by their surroundings: software graphics instead of a real graphics card, built-in functions replaced by scripts, a Web Worker or an iframe that reports other values than the page, a JavaScript engine that does not belong to the claimed browser, and a TLS handshake or HTTP/2 settings from a library rather than a browser.

How to read the result

• Not found: the check ran and saw nothing suspicious.
• Found: the sign is present. One sign is enough for many sites to show a captcha or block the visit.
• Signs we are still testing: shown as information and not counted yet. They become part of the verdict after testing on real devices.
• Verdict: this page looks only at automation. The full browser check also compares the browser with itself and with the network.

Further reading

• How headless browsers are detected
• navigator.webdriver explained
• How websites detect bots
• The API for checking visitors on your site

Frequently asked questions

I am a real person and the test says "found". Why?

The most common reasons are an extension that changes browser functions, an antivirus or corporate proxy that inspects HTTPS, or a virtual machine without a graphics card. Open the full check to see which sign it was.

Can I use this page to test my own automation?

Yes, testers and QA teams use it to see what their setups reveal. It shows the signs, not ways around other sites' protection.

Does the test use mouse movements?

No. It reads only what the browser and the connection report when the page loads.

Is anything stored?

No. The check runs once and is forgotten; only daily sums are counted, without the checks themselves.