Anti-Bot Detector: What Protects a Website (Free)

Anti-bot detector: enter a URL and see which bot protection the site uses (Cloudflare, Akamai, DataDome, HUMAN, Kasada and more) and whether a plain fetch can read it.

Ready. Enter a website, tick the box, press the button.

Anti-bot detector: enter a website, get the bot protection it shows and whether a plain fetch can read it. Background on bots: bot traffic.

What we check

We load the URL twice from one datacenter: with a Chrome user agent and with curl. We match the answers against vendor signs:

vendor what gives it away
Cloudflare cf-ray header, server: cloudflare, cf-mitigated: challenge header, __cf_bm / cf_clearance cookies, /cdn-cgi/challenge-platform/ script, the “Just a moment…” page
Akamai _abck, bm_sz, ak_bmsc cookies, server: AkamaiGHost, “Reference #” block page
DataDome datadome cookie, x-datadome header, js.datadome.co script, captcha-delivery.com challenge
HUMAN (PerimeterX) _px2 / _px3 / _pxhd / _pxvid cookies, px-cloud.net script, “Press & Hold” page
Kasada x-kpsdk-* headers, ips.js script
Imperva x-iinfo header, visid_incap_* / incap_ses_* / reese84 cookies, “Incapsula incident ID” page
AWS WAF x-amzn-waf-action header, aws-waf-token cookie, awswaf.com script
Sucuri, F5, Fastly, CloudFront their own headers and cookies
Captcha widgets reCAPTCHA, hCaptcha, Cloudflare Turnstile, Arkose Labs, GeeTest scripts

What comes back

  • Verdict: readable, challenged, blocked or no answer.
  • Protections seen: vendor, role (bot manager, firewall, CDN, captcha) and the proof we saw.
  • Two fetches: status, words of visible text, size for each user agent.

A challenge page counts only when the page is refused or nearly empty: an article that talks about Cloudflare is not a Cloudflare challenge.

Limits

  • One datacenter location, no JavaScript run. A page built in the browser shows few words even when nothing blocks it.
  • Protection that decides by network fingerprint, IP reputation or behavior leaves no sign in a plain fetch. Some sites pass one request and block the next.
  • A match means a vendor’s sign is present, not that every page is protected.
  • 10 checks per visitor per day, 6 per site per hour. Public addresses only.

Blocked on a site you need data from? Tell us the data you need.

Questions

What anti-bot is this site using?

Enter the address above. We load the page the way a script does and match headers, cookies, script addresses and challenge pages against known vendors. The table shows each match with what we saw.

How do I detect Cloudflare protection?

Look for the cf-ray response header, a server header saying cloudflare, the __cf_bm or cf_clearance cookies, and challenge pages titled Just a moment. Cloudflare marks its challenge responses with a cf-mitigated: challenge header.

Can a scraper read this site?

The result says what a plain fetch from one datacenter got: status, words of text, size. A page that answers 200 with real text is readable that way. A 403, a challenge page or an empty script shell is not. A real browser may still pass where a plain fetch fails, and the reverse.

Why does it say no protection when the site has some?

We read only what comes back in headers, cookies, script addresses and page text. Protection that scores visitors by network fingerprint or by behavior after the page loads leaves no trace there. No match means none seen.

Is checking a site's bot protection legal?

We make two ordinary page requests, like any visitor, and read what the site sends back. We do not solve challenges or bypass anything.

How often can I run it?

10 checks per visitor per day, and 6 per site per hour, so the tool cannot be used to hammer one site.