Glossary

AI Crawlers & Technical

ChatGPT-User

ChatGPT-User is the user agent OpenAI sends when ChatGPT fetches a web page live during a conversation, typically when a user asks it to browse, summarize or verify something. It acts on direct user requests rather than crawling systematically, making it a real-time fetcher distinct from GPTBot and OAI-SearchBot.

A fetcher, not a crawler

ChatGPT-User appears in your logs when a real person, mid-conversation, causes ChatGPT to open your page: pasting your URL, asking for a summary of your pricing, or requesting up-to-date details the model retrieves via browsing. Unlike GPTBot or OAI-SearchBot, it does not crawl proactively or build an index. Each hit maps to genuine human intent, which makes ChatGPT-User traffic one of the most commercially meaningful bot signals a site can receive.

Why these hits are a demand signal

A spike in ChatGPT-User fetches on a product or comparison page means users are actively discussing it inside ChatGPT, even if no click-through ever lands in your analytics. This is the invisible middle of the AI search funnel: research that happens entirely inside the chat interface. Blocking ChatGPT-User in robots.txt is self-defeating for almost everyone, because you are refusing a page view that a human explicitly requested; ChatGPT will simply tell the user it could not access your site, often pushing them to a competitor's page instead.

Tracking ChatGPT-User activity

Because these fetches execute no JavaScript, they are invisible to client-side analytics and only show up in server logs or edge-level tracking. Logging them per page reveals which content ChatGPT users actually pull into conversations, a different and often earlier signal than AI referral traffic. Geonimo's Cloudflare Worker integration records ChatGPT-User hits alongside other bot activity through its AI traffic analytics, letting you correlate in-conversation fetches with subsequent referral visits and citations.

Frequently asked questions

Should I block ChatGPT-User?

No, except for genuinely private content. Every ChatGPT-User fetch represents a human actively asking about your page inside ChatGPT. Blocking it means ChatGPT tells that user it cannot access your site, and the conversation moves on, frequently to a competitor whose pages remain accessible.

How is ChatGPT-User different from GPTBot?

GPTBot crawls the web systematically to gather model training data. ChatGPT-User fires only when an individual conversation triggers a live page fetch, such as a user pasting your URL or asking for current information. One is bulk collection; the other is on-demand retrieval tied to real user intent.

Why do ChatGPT-User visits not appear in Google Analytics?

The fetcher downloads your HTML but does not execute JavaScript, so analytics tags never fire. These visits exist only in server logs, CDN logs or edge workers that inspect requests before your page is served. Server-side tracking is the only way to count them.

Related terms

Last updated: 2026-06-11

Track this for your brand

Geonimo monitors how ChatGPT, Perplexity, Claude, Gemini and Google AI talk about your brand — and generates the content that gets you cited.

Get your free audit