// guide
How do you get cited by ChatGPT?
Updated: 2026-07-31
// short answer
To be cited by ChatGPT you need three things: crawler access for OAI-SearchBot and GPTBot in robots.txt, server-rendered pages with self-contained passages a model can quote without surrounding context, and independent sources that describe your brand the same way. Access and structure make you eligible; corroboration is what gets you named.
Start with crawler access, because it is binary
OpenAI documents separate user agents for separate jobs. GPTBot crawls for model training, OAI-SearchBot builds the search index behind ChatGPT's browsing, and ChatGPT-User fetches a page when a user's question triggers a live retrieval.
Blocking OAI-SearchBot removes you from the index ChatGPT searches, which is the one that matters for citations. Teams often block all three to protect content from training and then wonder why competitors are named instead. If the goal is visibility, the search and user agents need to be allowed at minimum.
- OAI-SearchBot — powers ChatGPT search; allow this one if you want to be cited.
- ChatGPT-User — fetches a page in response to a specific user request.
- GPTBot — crawls for training; blocking it does not remove you from search.
Render on the server
Retrieval fetches HTML and extracts text. It does not reliably execute JavaScript or wait for a client-side framework to hydrate. Content that only exists after hydration frequently reads as an empty page.
The practical check costs nothing: request the page with JavaScript disabled, or view the raw response, and confirm the claim you want quoted is present in the markup.
Write passages that survive being extracted
A model composing an answer takes fragments. A fragment that depends on the preceding three paragraphs to make sense is a poor candidate; one that states the answer completely on its own is a good one.
The reliable pattern is a heading phrased as the actual question, followed immediately by a compact answer that names the subject explicitly rather than saying 'it' or 'this approach', followed by the detail. Concrete numbers, dates and named sources make a passage more attractive to quote than the same claim stated vaguely.
Build corroboration off your own domain
For commercial questions — which vendor, which agency, which tool is best — assistants lean heavily on third-party sources: roundups, directories, community discussions, review platforms. A brand that exists only on its own website has nothing backing its claims.
This is the slowest part of the work and the part that most determines whether you are named. Getting into relevant listicles and directories, and being described consistently wherever you appear, does more than another round of on-page edits.
Make the entity unambiguous
Models need to resolve who you are before they can recommend you. Organization markup in JSON-LD, a consistent name and description everywhere, and a populated sameAs list pointing at real profiles all reduce ambiguity.
A sameAs array containing only your own homepage adds nothing — the point is to link the entity to independent places where it also appears.
Related questions
Does an llms.txt file make ChatGPT cite me?
There is no public confirmation that OpenAI consumes llms.txt. It is cheap to publish and helps any tool that does read it, but it should be treated as a small bet rather than a mechanism — crawler access and content structure are what demonstrably matter.
How quickly can a new page be cited?
Once crawled, a page can surface in live-retrieval answers within days. Answers drawn from the model's trained knowledge rather than live search lag much longer, which is why the same question can produce very different sourcing.
Why does ChatGPT cite my competitor and not me?
Usually one of three reasons: the search crawler is blocked or the page renders client-side, the relevant claim is not stated as a quotable passage, or the competitor is described by more independent sources. Check them in that order.
Should I block GPTBot to keep my content out of training?
That is a legitimate choice and it does not remove you from ChatGPT search, provided OAI-SearchBot stays allowed. Blocking the search and user agents is what costs you citations.