Does ChatGPT See My Website?
Short answer: maybe not. ChatGPT reaches websites with three different bots (GPTBot, OAI-SearchBot, ChatGPT-User), and any of them can be blocked even when Google sees your site just fine. Check below in 30 seconds — no signup required.
Free AI Readiness Check
See your GEO score before reading
Instant score · No email required · Checks ChatGPT, Gemini, Grok signals
How ChatGPT actually sees websites
Direct answer: not with one bot but with three, each doing a different job. GPTBot collects data for model training, OAI-SearchBot builds the search index behind ChatGPT Search, and ChatGPT-User fetches a page live when a user asks ChatGPT to open a specific site. Blocking one does not block the others — you need to check all three.
Collects content for training future model versions. It affects what ChatGPT 'knows' about you without searching — but the effect is slow: changes on your site only reach the model with the next training cycle, which means months. Blocking it is a deliberate choice, not a default: block it, and the model learns about you from other people's text.
The one that matters most for visibility right now. Its index decides whether your site shows up with a link in ChatGPT Search answers. If OAI-SearchBot is blocked or has never visited, you won't appear in ChatGPT's search answers — no matter what the model itself knows.
Fetches a page in real time when a user asks ChatGPT to open, summarize, or check a specific URL. It indexes nothing — it's a one-off visit. If it's blocked, ChatGPT can't open your site even when a user asks for it directly: 'I can't access that page.'
How to check it yourself in 2 minutes
Direct answer: two manual checks, no tools needed — open your robots.txt and look for these bots, and grep your server logs for the last few weeks. Both take minutes.
- 1. Open robots.txt
Go to yoursite.com/robots.txt and look for User-agent: GPTBot, OAI-SearchBot, ChatGPT-User. If any of them has Disallow: / underneath, that class of access is closed. If the bots aren't listed at all, the generic User-agent: * rule applies — check what it says.
- 2. Check your server logs
Search your access logs for the same user-agents: grep -i 'oai-searchbot\|gptbot\|chatgpt-user' access.log. If OAI-SearchBot hasn't visited in a month, you're not in ChatGPT's search index — and robots.txt isn't always the reason: starting September 15, 2026, Cloudflare blocks AI crawlers by default on new zones, so a bot can get a 403 even with a clean robots.txt. Check your CDN settings.
- 3. Ask ChatGPT itself
In search mode, ask: 'open yoursite.com and summarize the homepage.' If ChatGPT answers 'I can't access that page,' ChatGPT-User is blocked. If it opens the site but summarizes a nearly empty page, your content most likely only renders via JavaScript.
4 reasons ChatGPT might not see your site
Direct answer: it's almost always one of four things — AI bots blocked in robots.txt, content that only exists after JavaScript runs, missing structured data, or no brand mentions beyond your own site.
GPTBot, OAI-SearchBot, ClaudeBot, and PerplexityBot are separate user-agents, not Googlebot. If robots.txt was set up years ago or came from a hosting template, it can Disallow exactly these bots while regular Google crawling works fine.
Many AI crawlers don't execute JavaScript the way a user's browser does — they read the HTML the server returns. If all your text only appears after a React/Vue bundle loads, the crawler can end up with a nearly empty page.
Without Schema.org — and especially FAQPage markup — AI models see plain text with no explicit signal of 'here's a fact, here's the answer to a specific question.' FAQ Schema increases citation rate in AI responses by 41% based on our data — it's the fastest structural fix available.
AI models don't just trust a company's own site — they look for corroboration in mentions elsewhere. Brand mention correlation with AI citation frequency is higher than for classic backlinks. A site with no external footprint gets cited less often, even if it's technically flawless.
Nobody can promise your site will start showing up in ChatGPT answers within a week. Technical fixes (robots.txt, Schema.org) remove obstacles to visibility — necessary, but not sufficient. Actual citation depends on the query, competition, and brand signals, and takes weeks to months, not days.
What's fixable and what isn't
Direct answer: technical causes of invisibility — robots.txt, CDN-level blocks, JS-only rendering, missing markup — are fixable in hours or days. Reputational ones — brand mentions, competition for the query, what the model has already learned — take months of work, and no service can shortcut that.
- robots.txt and CDN settings — minutes: allow GPTBot, OAI-SearchBot and ChatGPT-User individually.
- Schema.org and FAQ markup — hours. Markup helps models extract facts, but on its own it doesn't decide visibility.
- Concrete facts on the page — days. Models measurably prefer pages with explicit prices, numbers and fresh dates, and skip vague wording with no facts. That's an editing job.
- llms.txt — minutes, but it's hygiene, not a hack: by itself it earns no citations.
- Brand mentions on external sites — models look for corroboration beyond your own site, and that footprint takes months to build.
- What the model has already learned: training data updates in cycles, and your changes won't get there immediately.
- Competition for the query — if strong sources already get cited for your topic, technical cleanliness only qualifies you for the race, it doesn't win it.
- Full client-side JS rendering — fixable, but it's a developer task (SSR or prerendering), not a checkbox in settings.
Platform guides
If you know what your site is built on, jump straight to the step-by-step guide for your platform:
Check your site for free
Causabi checks robots.txt, Schema.org, FAQ markup, and brand mentions — and generates ready-to-use files for your site.
Check your site