How to check if ChatGPT can see your website

Updated 18 October 2026 · AI Fix Shop

Check your site free: Run the free AI readiness checker (no sign-up, about 20 seconds).

Many owners now ask a simple question: when someone asks ChatGPT about my kind of business, can it even read my website? You can answer that yourself in about fifteen minutes. This guide shows the checks that matter, in the order that matters, and the mistakes that cause false alarms.

1. Know which bot actually matters

OpenAI runs more than one crawler, and they do different jobs. OAI-SearchBot is the one that finds and indexes pages so they can appear in ChatGPT search answers. GPTBot collects content for training models. ChatGPT-User fetches a page when a person in a chat asks ChatGPT to open it. Other AI tools follow the same pattern: Anthropic has Claude-SearchBot (search), ClaudeBot (training) and Claude-User; Perplexity has PerplexityBot and Perplexity-User.

So the first rule is: a GPTBot block is not proof that ChatGPT can't see you. It only means your content isn't used for training. For visibility in AI answers, the search bots are what count. Our GPTBot vs OAI-SearchBot guide explains the difference in detail.

2. Read your robots.txt

Open yoursite.com/robots.txt in a browser. Look for groups that name OAI-SearchBot, Claude-SearchBot or PerplexityBot followed by Disallow: /. Also check the catch-all group User-agent: *: if it says Disallow: /, every bot that respects robots.txt is shut out, AI search included. If the file doesn't exist (a 404), that's fine; it means nothing is blocked.

Many site builders and security plugins now add AI rules for you. If you didn't write the file yourself, read it anyway. Hosting dashboards and CDNs may also offer separate controls for search, training and agent bots, so check those settings match what you want.

3. Test what a bot receives

robots.txt is a request, not a wall. Firewalls and bot protection can still block a crawler that is allowed in. To test, fetch your homepage while identifying as the bot and compare it with a normal browser request:

curl -sI -A "Mozilla/5.0 (compatible; OAI-SearchBot/1.0; +https://openai.com/searchbot)" https://yoursite.com/
curl -sI -A "Mozilla/5.0 (Windows NT 10.0; Win64; x64)" https://yoursite.com/

A 200 for both is a good sign. A 403 or a challenge page only for the bot suggests a firewall rule. Be careful, though: real crawlers come from the AI company's published IP ranges, and your test comes from your own connection or a server. Some firewalls treat a spoofed bot name from an unknown IP as suspicious, so a block in your test is a signal to investigate, not proof. If the browser request is blocked too, the problem is your network, not AI access. Run the test twice, at different times, before drawing conclusions.

4. Check that the content is readable without JavaScript

Most crawlers read the raw HTML and do little or no JavaScript rendering. View the page source (not the inspector) and search for your main headline and a sentence from your services. If the source is mostly an empty <div id="root"> and script tags, AI tools may see a nearly blank page. Server-side rendering or pre-rendering fixes this.

5. Make it easy to understand who you are

Being reachable is step one; being understood is step two. A clear title and meta description, one H1, a readable About page and Organization schema help AI tools connect your site to your business name, location and official profiles.

6. Ask ChatGPT directly, with care

With search turned on, ask ChatGPT about your business by name and ask for a source link. If it cites your site, it can reach it. If it doesn't, that alone proves little: answers vary between runs and depend on what else is on the web. Treat it as a spot check, not a measurement.

Shortcut

Our free checker runs the robots.txt rules for the search bots, a bot-versus-browser fetch, a no-JavaScript text check, schema, titles, sitemap and llms.txt in one pass, and it labels training-only blocks as training-only so you don't chase the wrong problem.

FAQ

Does blocking GPTBot remove my site from ChatGPT?

No. GPTBot is OpenAI's training crawler. ChatGPT search uses OAI-SearchBot. Blocking GPTBot opts you out of training; blocking OAI-SearchBot is what keeps you out of ChatGPT search results.

Can I see AI crawler visits in my logs?

Yes. Search your server or CDN logs for user agents containing OAI-SearchBot, GPTBot, ChatGPT-User, Claude-SearchBot, ClaudeBot or PerplexityBot. Verify against the companies' published IP ranges, because anyone can fake a user agent.

My curl test with a bot user agent got blocked. Am I invisible?

Not necessarily. Real crawlers come from known IP ranges, and some firewalls block spoofed bot names from other IPs. Check robots.txt and your firewall rules, and repeat the test before concluding anything.

Check your site free: Run the free AI readiness checker (no sign-up, about 20 seconds).