How to get cited by ChatGPT
Last updated 2026-06-10
ChatGPT increasingly answers with a short list of cited sources instead of a page of links. Getting cited is concrete and mostly under your control: let OpenAI's crawlers in, be indexed in Bing (the index ChatGPT search actually retrieves from), make your answer extractable, and be trustworthy enough to attribute. Here's exactly how, in order of impact.
First, know OpenAI's three crawlers
OpenAI uses distinct user-agents for distinct jobs, and conflating them is the most common mistake:
- OAI-SearchBot — builds the search index ChatGPT pulls citations from. Allow this to be cited.
- ChatGPT-User — fetches a specific page live when a user asks ChatGPT to read it. Allow this for on-demand citations.
- GPTBot — gathers training data. Optional for citations — you can block it to opt out of training and still be cited.
Lever 1 — Allow the citation crawlers (robots.txt)
If OAI-SearchBot can't fetch the page, you cannot be cited — full stop. Make the policy explicit:
# Let ChatGPT find + cite you User-agent: OAI-SearchBot Allow: / User-agent: ChatGPT-User Allow: / # Optional: opt OUT of training, still get cited User-agent: GPTBot Disallow: /
Confirm your live score after editing with the validator.
Lever 2 — Be in the Bing index (the one ChatGPT reads)
This is the lever most guides skip, and it's the biggest one: ChatGPT search retrieves web results via Microsoft Bing's index— OpenAI's search partner — alongside its own OAI-SearchBot crawling. A page that ranks #1 on Google but is absent from Bing is largely invisible to ChatGPT. Three steps, ~10 minutes, free:
- Check your coverage: search
site:yourdomain.comon bing.com. Few or no results = you've found your problem. - Register at Bing Webmaster Tools — it can import your verified Google Search Console property in one click.
- Submit your sitemap there and fix anything its scan flags; Bing also honors IndexNow pings for same-day indexing of new URLs.
Lever 3 — Make the answer extractable
OpenAI's crawlers do not execute JavaScript, so client-only-rendered content is invisible to them — serve real HTML. Then lead each page with a tight, quote-ready answer (40–60 words) to the question it targets, use clear H2s, and add schema.org markup (Article, FAQPage, a Speakable block). A short FAQ mirroring real questions is one of the most reliably-cited formats, because each answer is a self-contained, attributable passage.
Lever 4 — Be trustworthy and current
Models prefer to cite claims they can corroborate and sources that look credible. Give pages a clear author/organization identity, an about/methodology page, and outbound citations to primary sources. Keep a visible and structured dateModified and refresh cornerstone pages — stale content gets cited less.
The checklist
- Allow OAI-SearchBot + ChatGPT-User in robots.txt.
- Register in Bing Webmaster Tools + submit your sitemap — ChatGPT search retrieves via Bing's index.
- Decide your GPTBot (training) policy separately.
- Serve real HTML — no JavaScript-gated content.
- Lead with a 40–60 word quote-ready answer; use clear H2s.
- Add Article + FAQPage + Speakable schema.
- Set + surface
dateModified; establish author identity. - Optional: publish an llms.txt — a cheap forward bet, though no provider has confirmed reading it yet.
- Verify with the validator.
FAQ
How do I get my website cited by ChatGPT?
Four things must be true. (1) ChatGPT's crawlers can reach you — allow OAI-SearchBot (search index) and ChatGPT-User (live fetches) in robots.txt. (2) You're indexed in Bing — ChatGPT search retrieves results via Bing's index, so register in Bing Webmaster Tools and submit your sitemap. (3) Your content is extractable — real server-rendered HTML, a tight quote-ready answer near the top, and schema.org markup. (4) You're trustworthy and current — clear author/organization identity and a recent dateModified. Being in the indexes is the prerequisite; blocking the crawlers removes you entirely.
Which crawler does ChatGPT use to cite sources?
ChatGPT search uses OAI-SearchBot to build the index it pulls citations from, and ChatGPT-User to fetch a specific page in real time when a user asks it to browse. GPTBot is separate — it gathers training data and does NOT need to be allowed for live citations. You can allow OAI-SearchBot + ChatGPT-User for citations while deciding on GPTBot separately.
Does blocking GPTBot stop ChatGPT from citing me?
No. Citations in ChatGPT search come from OAI-SearchBot's index and ChatGPT-User's live fetches, not from GPTBot (training). You can disallow GPTBot to opt out of training and still be cited, as long as OAI-SearchBot and ChatGPT-User are allowed.
Why isn't ChatGPT citing my page even though it ranks on Google?
Because Google's index is not the one ChatGPT reads. ChatGPT search is built on Bing's index plus OpenAI's own crawling — a page can rank #1 on Google while being absent or weak in Bing. Check Bing first: search site:yourdomain.com on bing.com, and register in Bing Webmaster Tools (you can import your verified Google Search Console property in one click). Then confirm OAI-SearchBot / ChatGPT-User are allowed in robots.txt and the answer is extractable near the top of the page.
Does Bing matter for ChatGPT citations?
Yes — it's the single biggest lever. ChatGPT search retrieves web results via Microsoft Bing's index (OpenAI's search partner) alongside OpenAI's own OAI-SearchBot crawling. If your site isn't indexed in Bing, ChatGPT mostly can't surface you no matter how good your content is. Register in Bing Webmaster Tools, submit your sitemap there, and verify coverage — it's free and takes about ten minutes.
Does an llms.txt file help with ChatGPT citations?
Honestly: there's no evidence it does today. No major AI provider has confirmed reading llms.txt, and Google has said publicly it doesn't use it. It costs nothing to publish and may pay off if adoption comes, so treat it as a cheap forward bet — never as a substitute for the levers that demonstrably work: Bing indexation, allowed crawlers, and extractable server-rendered HTML.
How long until ChatGPT starts citing me?
Once OAI-SearchBot has crawled and indexed your pages, you become eligible — re-crawl cadence is typically days to a few weeks. ChatGPT-User fetches are instant when a user asks it to read your URL. There's no instant switch for the index, but allowing the crawlers and shipping extractable content is the prerequisite everything else builds on.