OpenAI · Answers
ChatGPT-User
Someone typed a question, or pasted your URL, and ChatGPT went and got the page while they waited: one request, one human, one moment. OpenAI is emphatic that no scheduled sweep is involved and says this token is not used for crawling the web in an automatic fashion. Its scope covers Custom GPTs and GPT Actions as well as ordinary chat, so a third-party GPT calling out to your API can also arrive under this agent.
- Operated by
- OpenAI
- Purpose
- Answers
- robots.txt token
- ChatGPT-User
- Verification
- IP verified
Fetches a page in real time to answer a question someone just asked.
How to verify ChatGPT-User
OpenAI publishes the IP ranges ChatGPT-User crawls from, and we fetch that list on a schedule. A request claiming to be ChatGPT-User can therefore be checked against 219 published ranges: one that does not match is not this crawler.
Published ranges last confirmed by us on .
User agent
OpenAI publishes this user agent for ChatGPT-User. Match on the ChatGPT-User product token rather than the whole string: vendors revise the surrounding version and URL fragments without notice.
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; ChatGPT-User/1.0; +https://openai.com/botrobots.txt for ChatGPT-User
Hedged, and the hedge is the point: OpenAI states that because these actions are initiated by a user, robots.txt rules may not apply. Do not treat a disallow here as reliable. OpenAI redirects opt-out intent to OAI-SearchBot instead.
Block
User-agent: ChatGPT-User
Disallow: /Allow
User-agent: ChatGPT-User
Allow: /robots.txt is a request, not an enforcement mechanism. It is honoured by convention, and a crawler that ignores it is stopped at your edge, not in a text file.
What blocking ChatGPT-User costs you
A disallow here is unusual in that it may simply not be honoured, so the first honest thing to say is that blocking may not work at all. Where it does take effect (through a WAF, an IP rule, or an obliging fetch path), the person who lost out is someone who explicitly asked about your page and got nothing back, which is close to the highest-intent traffic on the open web. It costs you no search citations and no training exclusion, because both of those live on other tokens entirely.