How to Get Your Website Into ChatGPT Search
How to get your website on ChatGPT search: allow OAI-SearchBot, serve real HTML, get indexed on Bing, use IndexNow and check what ChatGPT can see.
To get your website on ChatGPT, you don't submit it anywhere. OpenAI's help article Searching the web with ChatGPT lists two conditions for a site to be eligible: allow OpenAI's search crawler, OAI-SearchBot, and make sure your host or CDN lets traffic from its published IP addresses through. In practice that comes down to four jobs: let OAI-SearchBot through robots.txt and your CDN, put the main content in the HTML, get your pages indexed on Bing, the web search provider OpenAI names, and tell Bing about changes with IndexNow. None of it guarantees a citation, and OpenAI says so: placement is not guaranteed.
This guide is about appearing in ChatGPT's search answers and citations. Building a website inside ChatGPT is a different feature, ChatGPT Sites, and pulling your Search Console data into a chat is covered in how to connect Google Search Console to ChatGPT.
How ChatGPT search finds and picks pages
When a prompt needs current information, ChatGPT searches the web and writes an answer that links to its sources. OpenAI's help article Searching the web with ChatGPT says ChatGPT search sometimes partners with other search providers, and that when it does, it typically rewrites your question into one or more targeted queries that it sends to them. OpenAI also runs its own crawler, OAI-SearchBot, and has an index of its own: its help center describes an offline web search option for some business workspaces that answers from OpenAI's indexed and cached web content instead of a live search, with no guarantee of covering the whole web.
So there are two routes into an answer, OpenAI's own crawl and a provider's search results, and a page can fail at a different stage on each. Keep the stages apart when something doesn't show up:
| Stage | What happens | What you control |
|---|---|---|
| Discovery | A search provider such as Bing learns the URL from links, a sitemap or IndexNow. OpenAI doesn't document how OAI-SearchBot finds URLs. | Internal and external links, a sitemap submitted to Bing, IndexNow |
| Crawling | OAI-SearchBot fetches the page for OpenAI; Bingbot fetches it for Bing. | robots.txt, CDN and firewall rules, server errors |
| Indexing | The page is stored in OpenAI's index or the provider's, where a search can retrieve it. | noindex, content present in the HTML, duplicates |
| Answering | For a given prompt, ChatGPT runs one or more searches, ranks the results, uses some pages and cites them. | How well the page answers the question; nobody controls this directly |
OpenAI's offline web search article also lists the common reasons a page is missing from its index or stale there: the site blocks crawling through robots.txt or similar controls, a CDN or bot protection blocks the crawler, the page needs a login, the content depends heavily on scripts, the page is new, rarely accessed or low-signal, or the site structure makes deeper pages harder to discover. The steps below cover the ones you control. The answering step is OpenAI's: ChatGPT ranks search results using multiple factors intended to help users find relevant, reliable information, and its transparency page names user intent, relevance and recency among them without publishing the details.
ChatGPT and Bing: what is documented
The short answer to whether ChatGPT uses Bing: in the cases that are documented, yes, but not only Bing. On May 23, 2023, Microsoft announced that “Bing is coming to ChatGPT as the default search experience”. When ChatGPT search launched on October 31, 2024, OpenAI wrote that any website or publisher can choose to appear, and that search leverages third-party search providers and content provided directly by OpenAI's partners, without naming the providers (Introducing ChatGPT search). As of October 2026, the help article on ChatGPT search for Enterprise and Edu says ChatGPT may share disassociated search queries with the Bing search engine to return web results. The consumer article, Searching the web with ChatGPT, refers to third-party search providers and points to Microsoft's and Shopify's privacy policies for how they process those queries.
A provider's ranking can carry over, too. OpenAI's transparency page says that when third-party search providers are used, the order of the sources in ChatGPT's sidebar is influenced by the provider's own ranking systems.
The honest reading: Bing is the web search provider OpenAI names, OpenAI also crawls on its own, and the mix behind any given answer isn't published. Being indexed on Bing covers the documented provider, and allowing OAI-SearchBot covers OpenAI's crawl. ChatGPT search SEO starts with doing both.
OpenAI's crawlers and what each one controls
OpenAI runs four user agents, listed in its Overview of OpenAI crawlers. Only one of them decides whether your pages can be in the ChatGPT search index:
| User agent | What it does | Effect on ChatGPT search |
|---|---|---|
OAI-SearchBot | Crawls pages to surface them in ChatGPT's search features. | Controls inclusion. Sites that block it aren't shown in ChatGPT search answers, though they can still appear as navigational links. |
GPTBot | Crawls content that may be used to train OpenAI's foundation models. | None. Blocking it signals that the content shouldn't be used for training. |
ChatGPT-User | Visits a page when a user's request in ChatGPT or a custom GPT needs it. Not an automatic crawler. | None. OpenAI says robots.txt rules may not apply to it, and it isn't used to decide search inclusion. |
OAI-AdsBot | Checks the safety of landing pages submitted as ads on ChatGPT. | None. It visits only pages submitted as ads. |
Each setting is independent: you can allow OAI-SearchBot and disallow GPTBot. If you allow both, OpenAI says it may use the results of one crawl for both purposes. OAI-SearchBot, GPTBot and ChatGPT-User covers each agent in detail.
How to get your website on ChatGPT, step by step
Getting indexed by ChatGPT means passing each stage above, so the steps come in the order things break. The first three decide whether OpenAI's crawler can read your pages, the next two cover Bing, and the last two are about what happens once you're eligible.
1. Allow OAI-SearchBot in robots.txt
Open https://www.example.com/robots.txt on your own domain. If no group names OAI-SearchBot and the * group doesn't disallow your pages, OAI-SearchBot can already crawl them: under the robots.txt standard, a crawler with no group of its own obeys the * group (RFC 9309). Look for the usual accidents: a Disallow: / under User-agent: * left over from staging, or a list of AI bots to block that includes OAI-SearchBot alongside GPTBot.
If you write rules per crawler, remember that a crawler with its own group ignores the * group: RFC 9309 has a crawler fall back to * only when no group matches it, and combines only the groups that do. Copy your shared Disallow lines into each named group:
# ChatGPT search: everything except private paths
User-agent: OAI-SearchBot
Disallow: /admin/
Disallow: /cart/
# Model training: opted out (no effect on ChatGPT search)
User-agent: GPTBot
Disallow: /
# Everyone else, Bingbot included
User-agent: *
Disallow: /admin/
Disallow: /cart/
Sitemap: https://www.example.com/sitemap.xmlThis lets OAI-SearchBot crawl everything except the admin and cart paths, opts the site out of model training, and leaves Bingbot and every other crawler on the * rules. Whether to allow GPTBot is a separate choice that doesn't change search. OpenAI's crawler overview says that after a robots.txt update, it can take about 24 hours for its systems to adjust search results. Which AI crawlers to allow in robots.txt covers GPTBot, ClaudeBot, Google-Extended and the others.
2. Let it through your CDN and firewall
robots.txt is a request; a firewall is a wall. OpenAI's search help says that to make a website eligible, you also need to confirm that your host or CDN allows traffic from OpenAI's published searchbot IP addresses. The ranges are at openai.com/searchbot.json, with separate files for the other three crawlers.
Bot protection is where this usually breaks. Cloudflare, for example, lists OAI-SearchBot in the AI Search category of its bot reference, and since July 1, 2026 every plan can block AI traffic by behavior, Search, Agent or Training, on all pages or only on pages that display ads (Cloudflare changelog). Under the new defaults, which apply to domains onboarding to Cloudflare from September 15, 2026, Search stays allowed. A rule that blocks Search, or a challenge page served to bots, keeps OAI-SearchBot out whatever robots.txt says. Check your provider's bot settings and your own WAF rules.
Then confirm in your access logs that OAI-SearchBot gets 200 responses. With the default nginx log format:
grep "OAI-SearchBot" /var/log/nginx/access.log \
| awk '{print $9, $7}' | sort | uniq -c | sort -rn | head -20That counts status codes per path. 403 or 503 lines point at a firewall or bot rule in front of your app. If the CDN blocks the bot, the request never reaches your server at all, so look in the CDN's firewall events too. Anyone can send that user agent string, so check the client IPs against searchbot.json before you trust the numbers.
3. Put the main content in the HTML
OpenAI doesn't document whether OAI-SearchBot runs JavaScript. Vercel, which measured crawler traffic on nextjs.org and across its network, reported in December 2024 that none of the major AI crawlers it observed rendered JavaScript, OpenAI's three included: they fetched JavaScript files without executing them (The rise of the AI crawler). OpenAI's offline web search article points the same way: it lists content that depends heavily on scripts or dynamic loading among the reasons a page can be missing or stale in OpenAI's index. A page whose text only appears after client-side rendering can look empty to a crawler that doesn't run it.
To see what a crawler that doesn't render gets, fetch the raw HTML and search it for a sentence from the page body:
curl -s https://www.example.com/blog/my-post | grep -c "a sentence from your main content"A count of 0 means the text isn't in the server response. Server-side rendering, static generation or prerendering fixes it; JavaScript SEO covers the options by framework. The same goes for titles, canonical tags and robots meta tags that only JavaScript adds.
4. Get indexed on Bing
Since Bing is the provider OpenAI names, a page missing from Bing's index is missing from that route into ChatGPT. Add the site to Bing Webmaster Tools. If it's already verified in Google Search Console, you can import it: Bing verifies imported sites automatically and brings their sitemaps across (Add and verify site).
Then submit your XML sitemap. Bing's Webmaster Guidelines list IndexNow, sitemaps, crawlable internal links and links from relevant sites as the ways it discovers URLs, and ask for sitemaps that list only canonical URLs, with lastmod where it applies. Bing calls lastmod a key signal and says it should reflect when the page content last changed (Keeping content discoverable with sitemaps in AI-powered search, July 2025). And make sure robots.txt doesn't block Bingbot. How to get your website indexed on Bing walks through it, and the Bing Webmaster Tools guide covers the reports.
5. Tell Bing about changes with IndexNow
IndexNow lets your site notify participating search engines that a URL was added, updated or deleted. A submission to one endpoint is shared with all of them (IndexNow FAQ), and as of October 2026 searchengines.json lists Bing, Yandex, Seznam.cz, Naver, Yep, the Internet Archive and Amazonbot. OpenAI isn't on the list, so IndexNow and ChatGPT have no direct connection: an IndexNow ping doesn't reach OAI-SearchBot. What it does is tell Bing that a page changed, without waiting for a recrawl. Bing says IndexNow helps keep information fresh across search and AI experiences, meaning its own.
After you host a key file on your domain, one POST covers up to 10,000 URLs (IndexNow documentation):
curl -X POST "https://api.indexnow.org/indexnow" \
-H "Content-Type: application/json; charset=utf-8" \
-d '{
"host": "www.example.com",
"key": "b6e3f0c94d2a4c1e9f7a8d5b3c2e1f00",
"urlList": [
"https://www.example.com/pricing",
"https://www.example.com/blog/new-post"
]
}'A 200 means the URLs were received; a 202 means they were received and key validation is pending. The FAQ adds that submitting a URL doesn't guarantee immediate indexing. Neither response is a crawl, an index entry or a citation, as IndexNow accepted is not indexed explains. What is IndexNow covers the key file and the protocol, Bing IndexNow the Bing Webmaster Tools side, and is IndexNow worth it the tradeoffs.
6. Use noindex and Disallow only where you mean it
A noindex left on from staging, a CMS setting or a template keeps pages out of Bing, and so off the Bing route into ChatGPT. Bing's Webmaster Guidelines say robots.txt controls crawl access, not indexing, and that noindex is the way to keep a URL out of Bing search, Copilot and grounding results. Check the HTML and the X-Robots-Tag response header of the pages you want found. The noindex tag covers finding and removing it.
On the OpenAI side, the documented control is robots.txt. Disallowing OAI-SearchBot keeps a site out of ChatGPT search answers, but OpenAI's crawler overview says such sites can still appear as navigational links, so a Disallow isn't a clean removal either. Use it only for paths you really want out of ChatGPT search.
7. Make the page worth citing
Access only makes a page eligible, and placement is not guaranteed. Microsoft, describing how its index supports AI answers in May 2026, said the unit of value for grounding shifts from documents to discrete, supportable facts with clear provenance, and that abstaining is a valid outcome when support is missing, stale or conflicting.
OpenAI doesn't publish its ranking factors, so take the general guidance for what it is: answer the question a page is about early and plainly, date and update facts that change, make the topic clear from the title and headings, and link to the page from related pages so crawlers reach it. How to get cited by ChatGPT goes into what makes a page citable, and AI search optimization covers the wider practice.
How to check whether ChatGPT can see your site
OpenAI's documentation describes no webmaster console for ChatGPT search, so you check from the outside. These four checks get you close:
- Server logs. OAI-SearchBot requests with
200responses on the pages you care about, from IPs insearchbot.json. If it never shows up, look upstream: robots.txt, the CDN, the firewall. - Referrals. Clicks on links in ChatGPT answers come from
chatgpt.com, so filter your analytics by that source and see which landing pages get them. - ChatGPT itself. With search on, ask a question your page answers, or ask for the page by name, and open Sources to see what was cited. Answers change with wording and over time, so treat this as a spot check, not a ranking.
- Bing Webmaster Tools. It shows whether Bing has your pages. Its AI Performance report, in public preview since February 2026, shows how your content appears across Microsoft Copilot, AI-generated summaries in Bing and select partner integrations. Microsoft doesn't name ChatGPT among them, so don't read it as a ChatGPT report.
What won't get you into ChatGPT search
- A submission form or a paid service. OpenAI's instructions for making a site eligible cover crawler access only, with no submission process. A service that offers to “index your site in ChatGPT” has the same documented levers as you, the ones above, and no one can guarantee a citation.
- Allowing GPTBot. It's the training crawler. Its robots.txt setting is independent from OAI-SearchBot's, so allowing it doesn't put you in search, and blocking it doesn't take you out.
- An llms.txt file. OpenAI's eligibility instructions and its crawler overview name OAI-SearchBot access as the control for search, not llms.txt. llms.txt explains what it is and where it can still help.
- Google Search Console on its own. Request indexing asks Google to crawl a URL, and OpenAI's documentation names no link between Google's index and ChatGPT search. Getting indexed on Google still matters for Google Search and its AI features: see how to get Google to index your website.
Claude, Perplexity and other AI search engines
Anthropic and Perplexity also split their crawlers by job. Cloudflare's bot reference lists Claude-SearchBot and PerplexityBot under AI Search, ClaudeBot under AI Crawler, and Claude-User and Perplexity-User under AI Assistant. The same checks apply: find the search crawler, allow it, let it past your firewall and serve it HTML.
- Claude: ClaudeBot, Claude-SearchBot and Claude-User
- Perplexity: Perplexity SEO
- Google: do you need to be indexed to appear in AI Overviews
- All of them in one file: GPTBot and other AI crawlers in robots.txt
- The practice as a whole: AI search engine optimization and getting cited by ChatGPT
How IndexCyborg handles the Bing side
IndexCyborg automates the parts of this that go through Bing and the other IndexNow engines. Once you verify that you own the domain (DNS TXT record or HTML file) and host an IndexNow key file, it reads your sitemaps, detects new and changed pages on every rescan (every 24 hours by default) and notifies them through api.indexnow.org, up to 1,000 URLs per call, logging each response word for word. Its technical checks cover HTTP status, redirect chains, noindex in meta robots or X-Robots-Tag, canonical tags, titles and descriptions; each issue comes with evidence and is resolved automatically once you fix it.
It doesn't notify OpenAI, which isn't an IndexNow participant, and it doesn't track ChatGPT citations or test your robots.txt and firewall against OAI-SearchBot: its robots.txt check is evaluated for Googlebot. Use the log and referral checks above for that. On the Google side, it monitors index status with the URL Inspection API and requests indexing through its Chrome extension, in your own Search Console session. More on the features page.
Frequently asked questions
Can I submit my website to ChatGPT?
No. OpenAI has no submission form for ChatGPT search; its help center says a site becomes eligible when OAI-SearchBot can crawl it and the host or CDN lets OpenAI's search crawler IP addresses through. Also get your pages indexed on Bing, the web search provider OpenAI names for Enterprise and Edu.
Does blocking GPTBot remove my site from ChatGPT search?
No. GPTBot collects content that may be used to train OpenAI's models, and OpenAI treats its robots.txt setting as independent from OAI-SearchBot's. You can disallow GPTBot and still allow OAI-SearchBot, the crawler that controls search inclusion.
How long does it take for a website to show up in ChatGPT?
OpenAI publishes no timeline for crawling or including new pages and says there is no refresh SLA for a specific URL in its index. The one figure it gives is that a robots.txt change can take about 24 hours to be reflected in search. On the Bing side, IndexNow tells Bing about a new or changed URL right away, but Bing still decides when to crawl it and whether to index it.
Do I have to pay to appear in ChatGPT search?
No. OpenAI's eligibility conditions for search are about crawler access, and it describes no fee or paid inclusion for search answers. Ads in ChatGPT are a separate product: OpenAI's OAI-AdsBot visits only landing pages submitted as ads, to check them.
Can ChatGPT read my page if OAI-SearchBot is blocked?
Sometimes. When a user asks ChatGPT to open a page, the visit comes from ChatGPT-User, which isn't an automatic crawler, and OpenAI says robots.txt rules may not apply to it. That doesn't make the page eligible for search answers: OpenAI uses OAI-SearchBot, not ChatGPT-User, to decide search inclusion.