Free · runs in your browser · no signup
How much of your AI-bot traffic is fake?
Anyone can send “GPTBot” or “ClaudeBot” in a request. Drop in your access log and we check every AI-bot request against the official IP ranges published by 16 AI and search crawlers, so you see which bots really read your pages.
Drop your access log here
nginx or Apache, plain or .gz, any size. Processed in your browser: the file never leaves your device.
…or paste log lines
User agents lie
The User-Agent header is free text. On our own servers, at least 43% of requests claiming to be AI bots were attack probes for files like .env, which no real crawler requests. Read the study
IP ranges don't
OpenAI, Anthropic, Perplexity, Google, Microsoft, Apple, Common Crawl, DuckDuckGo and Mistral publish the IP ranges their bots use. We download every list daily and match each request against it.
Know what AI reads
See which pages verified crawlers fetch, which AI assistants send you visitors, and which bots can't be verified at all because their operator publishes no list.
AI bots we verify
Full directory →Verify every AI-bot visit, continuously
On WordPress? Our free plugin checks every AI-bot visit as it happens and shows real vs fake per bot, the pages AI reads and the visitors AI assistants send you. Get the WordPress plugin. Not on WordPress? Join the beta for daily reports and alerts.
Join the beta listQuestions
Can a bot fake being GPTBot or ClaudeBot?
Yes. The User-Agent header is just text that any client can set. Vulnerability scanners routinely send AI-crawler user agents. The reliable check is the source IP: OpenAI, Anthropic, Perplexity, Google, Microsoft, Apple and others publish the IP ranges their crawlers use.
How do I verify GPTBot?
Check that the request's IP address falls inside OpenAI's published list (openai.com/gptbot.json). The same works for OAI-SearchBot (searchbot.json) and ChatGPT-User (chatgpt-user.json). This checker does that for every line of your log.
Does Anthropic publish ClaudeBot IP addresses?
Yes. Anthropic publishes the addresses its bots use at claude.com/crawling/bots.json. Many older articles still say it doesn't.
Is Google-Extended a crawler?
No. Google says Google-Extended has no separate HTTP user agent: it is only a robots.txt token. Any request with “Google-Extended” in its User-Agent is fake.
Is my log uploaded anywhere?
No. The checker runs entirely in your browser. Your log file is read locally and never sent to our server.
IP lists last refreshed 2026-10-21.