AEO Log Analyzer

A free client-side tool that parses your server access log and identifies which AI crawlers actually hit your site. It matches the User-Agent of every request against 18 known crawler patterns, then verifies the source IP against the operator's published IP ranges — Googlebot, Bingbot, Applebot, GPTBot, OAI-SearchBot, ChatGPT-User and Amazonbot all publish ranges, so hits are shown as verified or unverified. The tool also surfaces status codes, rate-limit responses (401, 403, 429, 503), first and last seen timestamps, and top requested paths per bot.

Download your server access log for the period you want to analyze (Apache combined, Nginx combined, Nginx JSON, or Cloudflare CSV format), then drop the file into the AEO Log Analyzer. The tool parses each line, identifies crawler User-Agents, and cross-references the source IP against the operators' published IP ranges. Requests where both the UA and IP match are marked verified. Requests where the UA claims a known crawler but the IP does not match any published range are marked unverified — usually spoofed scrapers rather than the real bot.

GPTBot is verified against OpenAI's published IP ranges. If a hit is shown as unverified, the request's source IP is not in OpenAI's published range for GPTBot — even though the User-Agent claims to be GPTBot. This is usually a spoofed scraper, not the real OpenAI crawler. Any tool or person can set the User-Agent string to "GPTBot", but only OpenAI's infrastructure can send requests from OpenAI's IP ranges. The unverified count in the report shows you how many hits claimed to be a given bot but came from a non-matching source.

HTTP 429 means "Too Many Requests" — the server (or a CDN/WAF in front of it) is rate-limiting the crawler because it sent too many requests in a short window. For AI crawlers this is important: if a verified crawler is being rate-limited, it may reduce its crawl rate for your site, delay re-visits, or skip pages entirely. The AEO Log Analyzer flags 429 responses per bot so you can see whether your rate limits are blocking legitimate crawlers. Other suspicious statuses include 401 (Unauthorized), 403 (Forbidden), and 503 (Service Unavailable).

No. The tool parses your log file entirely in your browser. Nothing is uploaded to any server. The only network request the tool makes is to fetch the operators' published IP ranges — a public, static list that contains no information about your site or your logs. Your IP addresses, URL paths, referrers, and User-Agent data never leave your device.

Yes. The tool auto-detects and parses four common access log formats: Apache combined (the default on most Apache setups), Nginx combined (the default on most Nginx setups), Nginx JSON (if you have configured Nginx with a JSON log format), and Cloudflare CSV (exported from the Cloudflare dashboard). The format is detected automatically from the first few lines of the file. If you paste a log in an unsupported format, the tool will tell you rather than silently parsing it wrong.