How to get your site into ChatGPT search
By Dhrumil Barot, Founder
Allow OAI-SearchBot, keep pages crawlable, get indexed in Bing and Google, and check readiness. Sourced from OpenAI’s own documentation, with the unknowns stated.
Published . Updated .
To appear in ChatGPT search, let OpenAI’s search crawler, OAI-SearchBot, fetch your pages, keep the pages readable and indexable, and get them indexed in Bing and Google. There is no submission form we can find. Below is what OpenAI documents, what third parties report, and what nobody outside OpenAI can promise.
Key takeaways
- OpenAI says sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links.
- Allow OAI-SearchBot in robots.txt, and allow OpenAI’s published IP ranges at your firewall or CDN.
- OAI-SearchBot (search) and GPTBot (training) are independent, so you can allow one and block the other.
- Bing’s index is widely reported to feed ChatGPT search, but OpenAI says it also uses other providers. Treat Bing as helpful, not decisive.
- No setting guarantees a citation. You can only remove the blockers and make the page easy to read.
Step 1: Allow OAI-SearchBot in robots.txt
OpenAI’s crawler documentation says OAI-SearchBot is used to surface websites in ChatGPT’s search features, and recommends allowing it in your robots.txt file. A rule that allows it looks like this, and you can leave it out entirely if your robots.txt does not block it already:
User-agent: OAI-SearchBot
Allow: /
Then check every other layer that can block it: a blanket Disallow for all crawlers, a bot-fighting setting at your CDN, or a firewall rule. OpenAI publishes the IP ranges of OAI-SearchBot at openai.com/searchbot.json so you can allow them. Per OpenAI, changes can take about 24 hours to take effect. Source: OpenAI crawler documentation. Our robots.txt checker shows which rule applies to a path for a crawler you select, and our blog covers why AI crawlers get blocked even when robots.txt is open.
Step 2: Decide separately about training
GPTBot is OpenAI’s training crawler. Per the same documentation, disallowing GPTBot signals that content should not be used to train its generative AI foundation models, and that setting is independent of OAI-SearchBot. ChatGPT-User is different again: it acts on a user’s request inside ChatGPT, is not used for automatic crawling, and robots.txt rules may not apply to it.
| User agent | What OpenAI says it is for | What to do for search visibility |
|---|---|---|
| OAI-SearchBot | Surfacing sites in ChatGPT search features | Allow it |
| GPTBot | Crawling content that may be used for training | Your choice; independent of search |
| ChatGPT-User | Certain user actions in ChatGPT and custom GPTs; not automatic crawling | Cannot be managed reliably through robots.txt per OpenAI |
Step 3: Make sure the page can be read and is indexed
If a page is not available to be crawled and read, there is nothing to cite. Check the basics: the page returns 200 without a login, is not noindex, has the right canonical, and has its main text in the HTML it serves rather than only after JavaScript runs. Our URL health checker checks the first three. If the page is new, send it to the search engines so it is discovered sooner; how to help AI search find your website covers the wider picture.
Step 4: Get indexed in Bing, and in Google
Several third-party sources, including Yoast, report that ChatGPT search uses Bing’s index, and Yoast concludes that a site missing from Bing is unlikely to appear. OpenAI’s position, as quoted by Semrush, is that Bing is the only third-party search provider for Enterprise and Edu workspaces, while OpenAI also says ChatGPT search sometimes partners with other search providers. Semrush also reports tests in which ChatGPT cited a page reachable only through Google. The sources do not fully agree, so cover both: submit your site to Bing and make sure Google has indexed it. Sources: Yoast (November 2024) and Semrush (August 2025).
Step 5: Make pages easy to quote
Nobody outside OpenAI knows exactly how it chooses sources, so be wary of anyone selling a formula. What you can do is ordinary good practice: a clear title, headings that match real questions, concise answers near the top, a named author, a visible date and accurate facts. These help readers as much as crawlers.
Step 6: Check readiness for free
Instead of checking each of the points above by hand, run the free AI search readiness checker. It grades a page from A to F and checks, among other things, whether robots.txt and your firewall allow OAI-SearchBot and other AI crawlers, whether the text is in the initial HTML, structured data, and whether llms.txt exists. It follows fixed rules rather than using AI, so the same page gets the same score, and it needs no account.
Let IndexLinks send the pages and show who came
Once the blockers are gone, discovery is the next gap. IndexLinks sends new and updated pages to Bing, Yandex, Naver, Seznam and Yep, to Google through your Search Console, and to Baidu, and makes them available to GPTBot, ClaudeBot and PerplexityBot through its Discover hub. It then shows a verified receipt for each crawler visit, so you can see whether a crawler arrived. It cannot make ChatGPT cite you, and it does not claim to. See plans or run the free readiness check first.
Frequently asked questions
Is there a way to submit my site to ChatGPT?
We could not find a submission form in OpenAI’s documentation. What it does document is a crawler, OAI-SearchBot, and says that to appear in ChatGPT search results you should allow it in robots.txt.
Should I block GPTBot?
That is separate from search. OpenAI says each setting is independent: you can allow OAI-SearchBot to appear in search results and disallow GPTBot to signal your content should not be used to train its models.
How long after changing robots.txt does ChatGPT notice?
OpenAI says it can take about 24 hours from a robots.txt update for its systems to adjust.
Does ChatGPT search use Bing?
Reports and Microsoft’s announcement say Bing is a search provider for ChatGPT, and OpenAI says Bing is the only third-party search provider for Enterprise and Edu workspaces. OpenAI also says ChatGPT search sometimes partners with other providers, and tests reported in 2025 suggest Google can be involved. Getting indexed in Bing is sensible, but it is not a guarantee.
Can I guarantee ChatGPT will cite my site?
No. Allowing access only makes a page eligible. Whether it is chosen for an answer depends on the question and the sources available, and nobody outside OpenAI can promise it.
Does ChatGPT-User follow robots.txt?
OpenAI says ChatGPT-User handles actions started by a user and is not used for automatic crawling, so robots.txt rules may not apply to it. Use OAI-SearchBot to manage search opt-outs.
Use this information as a starting point, then verify the page-level evidence available to you.