Free tool
Paste an IP address and user agent from your server logs. We check them against the operator's published IP list or DNS records.
Free. No account needed. The IP you paste is used for this check only and isn't stored.
Try an example
A Googlebot request from an address in Google's published list.A GPTBot request from an address in OpenAI's published GPTBot list.A request that claims to be GPTBot from an address OpenAI doesn't list.A Bytespider request. ByteDance doesn't publish a way to verify it.A request from an ordinary Chrome browser.The user agent is a text label the client chooses. Any script can send GPTBot or Googlebot in its requests, and scrapers often do, because some sites let known crawlers through.
Real verification uses something an impostor can't fake: the network the request came from. The operator says which addresses or DNS names belong to its crawler, and the request either comes from one of them or it doesn't.
Published IP lists. OpenAI, Anthropic, Perplexity, Google, Bing, Apple and others publish the address ranges their crawlers use. We fetch each list from the operator, keep it for a day, and show you which list and range matched and how old the list is.
Reverse DNS with a forward check. Google, Bing, Apple, Yandex and Baidu document this method. A lookup on the IP must return a hostname on the operator's domain (for example crawl-66-249-66-1.googlebot.com), and a lookup on that hostname must return the same IP. Both steps must pass.
Signed requests. A few operators sign their requests. The signature travels in the request headers, so it can't be checked from a log line, and this tester doesn't claim to.
The AI Crawler Directory shows which method each crawler supports.
robots.txt won't help. An impostor isn't the crawler it names, so it has no reason to follow that crawler's rules, and blocking the token would only stop the real one.
Block the IP or range at your CDN or firewall instead. If it keeps happening, measure how often: one request is noise, thousands a day is a scraper.
No. It's used for this check and then discarded. It isn't logged, put in the page address or shared.
Some operators say fetches a user asked for may not follow robots.txt. Check the crawler's page in the AI Crawler Directory.
Not from a log line, because the signature is in request headers.
Your server sees the CDN. Use the original client IP from the CDN's header (for example CF-Connecting-IP or X-Forwarded-For).
Every major AI crawler: who runs it, what it collects, how to block it.
Browse crawlersSee which AI crawlers your robots.txt allows or blocks.
Check robots.txtPaste referrer URLs and see which visits came from AI, search or social.
Classify referrersConnect your site today and see your first verified AI crawler by tomorrow.