Is this IP really that AI bot?
When your server log shows a User-Agent such as "GPTBot" or "ClaudeBot", this checks whether the IP address of that request is really inside the IP ranges that the bot's operator publishes.
The four answers
verified: This IP address is inside the IP ranges that the operator of the bot named by the User-Agent publishes.ip_not_in_feed: The User-Agent names a registered bot, but this IP address is not in the ranges its operator publishes. The request may only be borrowing the name.no_feed: The User-Agent names a registered bot, but its operator publishes no IP ranges (or the snapshot has none), so this IP address cannot be checked. It does not mean the request is fake.unknown_ua: This User-Agent matches none of the bots in the registry. It cannot be checked.
How it works
- The User-Agent is matched against the bot registry (bot names and UA patterns); then the IP address (IPv4 or IPv6) is checked against the snapshot of the IP ranges that bot's operator publishes.
- The data is only what this Worker bundled at build time; no outside request is made when you check. The answer names the bot, the feed and its URL, the snapshot date and the registry version.
- "verified" only means the IP address is inside the operator's published ranges (it does not tell which of the operator's bots). The snapshot goes stale between updates.
- API:
GET /en/check/ai-bot-ip?ip=…&ua=…(JSON; HTML for a browser, Markdown forAccept: text/markdown). The MCP tool isverify_ai_bot(ip,user_agent).
Checking a whole site
For whether bots may fetch your site, use the free scan.