ChatGPT-User user agent and robots.txt rules

ChatGPT-User fetches pages for actions people start in ChatGPT and custom GPTs, such as opening a link to answer a question. It does not crawl the web on its own.

User-triggered May skip robots.txt OpenAI

User agent

As documented by OpenAI. Version numbers change; match on the token.

User agent
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; ChatGPT-User/1.0; +https://openai.com/bot

robots.txt rules for ChatGPT-User

Paste one block into the robots.txt at the root of your domain. Changes can take up to a day to be read.

Block ChatGPT-User

Disallows every path for ChatGPT-User. That asks OpenAI not to fetch your pages when its users ask, though OpenAI says user-triggered fetches may not always follow robots.txt.

robots.txt
User-agent: ChatGPT-User
Disallow: /

Allow ChatGPT-User

Lets ChatGPT-User fetch every path, whatever “User-agent: *” says.

robots.txt
User-agent: ChatGPT-User
Allow: /

Good to know

OpenAI says robots.txt rules may not apply to it, because a user started the request. It doesn’t decide what appears in ChatGPT search.

ChatGPT-User FAQ

What is ChatGPT-User?

ChatGPT-User fetches pages for actions people start in ChatGPT and custom GPTs, such as opening a link to answer a question. It does not crawl the web on its own.

How do I block ChatGPT-User?

Add “User-agent: ChatGPT-User” and “Disallow: /” to your robots.txt. That asks OpenAI not to fetch your pages when its users ask, though OpenAI says user-triggered fetches may not always follow robots.txt.

Does ChatGPT-User respect robots.txt?

OpenAI says that because a person starts these requests, robots.txt rules may not apply to ChatGPT-User.

What is the ChatGPT-User user agent?

OpenAI documents it as “Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; ChatGPT-User/1.0; +https://openai.com/bot”. Version numbers change, so match on “ChatGPT-User”.

How can I verify ChatGPT-User requests?

Match the request IP against OpenAI’s published list (https://openai.com/chatgpt-user.json). The user agent alone can be faked.

Source: developers.openai.com, checked Nov 17, 2026. AI Crawler Directory

Files are the start. Answers are the result.

Your buyers’ questions on eight AI engines, on your schedule, with history. Free for 7 days.