AI crawler user agent checker

Copy the User-Agent value from a trusted server or edge log. Paste that field alone, rather than a complete log line containing addresses, cookies or other request data. The lookup runs entirely in your browser and does not send the pasted string to SearchSeal or another service.

A match shows the crawler token, owner, purpose and verification guidance. The checker recognizes token boundaries and ignores token names that appear only inside a URL. If several recognized identities appear, it shows them separately so you can investigate the original request instead of silently choosing one.

Search indexing, training and user-requested fetching are different activities. A crawler that helps build a search index serves a different purpose from a client retrieving a page after someone asks a question. Read the purpose in the result before deciding whether to allow or restrict that activity.

The reference covers fourteen tokens, including the OpenAI, Anthropic and Perplexity crawler families, plus Google-Extended, Applebot-Extended, Common Crawl, ByteDance, Meta and Amazon entries. Where current owner guidance could not be confirmed, the result says so rather than assigning a precise policy or verification method.

A user-agent string identifies a claimed client and can be copied by someone else. To verify a request, use the source IP recorded by your trusted edge or server and compare it with the current method published by the owner. Several owners publish IP ranges; others also document reverse DNS checks.

When the documented method uses reverse DNS, resolve the source address to a hostname, check the exact allowed domain suffix and resolve that hostname forward to confirm the original address. Follow the owner's complete procedure. Do not create a broad firewall exception based solely on the text returned by this lookup.

Google-Extended and Applebot-Extended control certain uses of content through robots.txt. They are not separate HTTP crawler identities. If you paste either token, the result explains that distinction and points to the owner guidance for the actual requests you might observe.

An unknown result means the string did not match this catalog. Keep it unidentified while investigating. Check whether the log was truncated and whether you copied the complete field. A familiar browser name elsewhere in the string does not establish which service made the request.

Yes. A token such as OAI-SearchBot is enough to open its reference entry. Use the complete field when investigating a real request, because it preserves version information and reveals whether several identities appear together. The checker matches names without depending on a particular version number, while keeping similarly named but unrelated strings separate.

Once you have a verified identity, review the path, request time and HTTP status. These fields tell you what the client asked for and how the site responded. Keep the verification method beside that record so a future reviewer can understand the classification.

Use the robots.txt AI crawler checker to inspect published rules for the same path. Read the crawler request guide for the wider log investigation. If the question concerns what answers say about your business, explore AI brand monitoring, which uses saved answer evidence rather than request traffic.