AI Visibility Checker: check if ChatGPT and AI crawlers can read your website

Check if ChatGPT, Claude, Perplexity, Gemini and Bing can read your website, in seconds. Paste one or more URLs and get a score out of 100, the verdict for each AI crawler in your robots.txt, and one sentence per fix. Free, no AI calls, no login to any third party.

AI assistants can only recommend what they can read. A site can rank on Google and still be invisible to AI because its robots.txt blocks the AI search bots, because the content only appears after JavaScript runs, or because there is no structured data. This Actor finds those problems with plain HTTP requests.

What it checks

Check	Points	What it looks at
AI search and assistant crawlers in robots.txt	25	OAI-SearchBot, ChatGPT-User, Claude-SearchBot, Claude-User, PerplexityBot, Perplexity-User, Googlebot, Bingbot: allowed or blocked for this exact page, with the matching rule
AI training crawlers in robots.txt	5	GPTBot, ClaudeBot, Google-Extended, Applebot-Extended (blocking them is a valid choice, so it weighs little)
Indexable	5	No noindex in the meta robots tag or the X-Robots-Tag header
Text readable without JavaScript	20	Word count in the raw HTML, the way a crawler that does not run JavaScript sees it; flags pages rendered in the browser (React, Vue, Angular, Next.js shells)
Structured data (JSON-LD)	15	JSON-LD blocks, parse errors, schema.org types found (Organization, Product, SoftwareApplication, FAQPage, Article...)
Title tag	5	Present, 10 to 70 characters
Meta description	5	Present, 50 to 170 characters
Canonical	5	Points to the page itself, to another URL, or missing
Language	3	<html lang> declared
XML sitemap	7	Declared in robots.txt or found at /sitemap.xml, number of <loc> entries
llms.txt	5	/llms.txt exists as a text file (not an HTML fallback), its size, plus /llms-full.txt

robots.txt is parsed like the standard says (RFC 9309): the most specific user-agent group wins, then the longest matching rule, Allow wins ties, * and $ wildcards are supported. robots.txt, llms.txt and the sitemap are fetched once per domain, even if you check many pages of the same site.

Grades: A = 90 and up, B = 75 and up, C = 60 and up, D = 40 and up, F below 40.

How to use

1. Click Try for free.
2. Paste your URLs in Website URLs to check, one per line (home page, pricing page, a product page, a blog post). Up to 100 per run.
3. Click Start. Each URL takes a few seconds.
4. Open the Output tab: the Overview view shows the score and fixes, All checks shows every detail. The SUMMARY record in the key-value store gives the average score and the top 3 fixes per URL.

You can also call it from the API, a schedule (to watch your robots.txt after each deploy), Make, Zapier, n8n or any MCP client.

Input example

{
  "urls": ["https://example.com", "https://example.com/pricing"],
  "timeoutSecs": 15
}

Output example

Real result for https://howseen.ai (checks list shortened to the first 4 of 11, bots list shortened to 2 of 12):

{
  "url": "https://howseen.ai",
  "finalUrl": "https://howseen.ai/",
  "statusCode": 200,
  "score": 100,
  "grade": "A",
  "checks": [
    {
      "id": "robots_ai_search",
      "label": "AI search and assistant crawlers allowed in robots.txt",
      "status": "pass",
      "points": 25,
      "maxPoints": 25,
      "detail": "All 8 search and assistant crawlers are allowed.",
      "recommendation": null
    },
    {
      "id": "robots_ai_training",
      "label": "AI training crawlers allowed in robots.txt",
      "status": "pass",
      "points": 5,
      "maxPoints": 5,
      "detail": "All training crawlers are allowed.",
      "recommendation": null
    },
    {
      "id": "indexable",
      "label": "Page is indexable (no noindex)",
      "status": "pass",
      "points": 5,
      "maxPoints": 5,
      "detail": "No noindex directive found.",
      "recommendation": null
    },
    {
      "id": "server_text",
      "label": "Text readable without JavaScript",
      "status": "pass",
      "points": 20,
      "maxPoints": 20,
      "detail": "2976 words in the raw HTML.",
      "recommendation": null
    }
  ],
  "recommendations": [],
  "robotsTxt": {
    "url": "https://howseen.ai/robots.txt",
    "found": true,
    "bots": {
      "OAI-SearchBot": {
        "company": "OpenAI",
        "purpose": "search",
        "allowed": true,
        "matchedGroup": "oai-searchbot",
        "rule": "Allow: /",
        "note": "ChatGPT search results"
      },
      "Googlebot": {
        "company": "Google",
        "purpose": "search",
        "allowed": true,
        "matchedGroup": "*",
        "rule": "Allow: /",
        "note": "Google Search, AI Overviews and AI Mode"
      }
    }
  },
  "llmsTxt": {
    "url": "https://howseen.ai/llms.txt",
    "found": true,
    "status": 200,
    "bytes": 10132,
    "llmsFullTxt": {
      "url": "https://howseen.ai/llms-full.txt",
      "found": true,
      "bytes": 5256
    }
  },
  "jsonLd": {
    "blocks": 5,
    "invalid": 0,
    "types": [
      "AggregateOffer",
      "Answer",
      "FAQPage",
      "Organization",
      "Person",
      "Question",
      "SoftwareApplication",
      "WebPage",
      "WebSite"
    ]
  },
  "content": {
    "wordCount": 2976,
    "likelyClientRendered": false,
    "frameworkHints": [],
    "htmlBytes": 298236
  },
  "canonical": {
    "href": "https://howseen.ai/",
    "pointsToSelf": true
  },
  "title": "Howseen AI: Track Your Brand in ChatGPT, Gemini & Perplexity",
  "lang": "en",
  "sitemap": {
    "found": true,
    "url": "https://howseen.ai/sitemap.xml",
    "source": "robots.txt",
    "urlCount": 4,
    "declaredInRobots": true
  },
  "howseen": "This checks if AI can read your site. To see if ChatGPT, Gemini, Perplexity and Google AI actually recommend your brand, try Howseen: https://howseen.ai/?utm_source=apify&utm_medium=actor&utm_campaign=ai-visibility-checker"
}

When something is wrong, recommendations lists one sentence per fix. Real recommendations returned for a JavaScript single-page app we tested (score 60/100, 1 word in the raw HTML):

• Most AI crawlers do not run JavaScript: render the main content on the server (SSR or static HTML) so they see it.
• Add JSON-LD (Organization plus Product, Service or FAQPage) in the server HTML so AI gets your facts in a machine-readable form.
• Add a /llms.txt that sums up who you are and links your key pages in Markdown (proposed standard, llmstxt.org).

FAQ

Does this tell me if ChatGPT recommends my brand? No. It checks whether AI crawlers are allowed in and whether they can read your page. Being readable is the first step; being recommended depends on what the AI finds about you across the web. To see if ChatGPT, Gemini, Perplexity and Google AI actually recommend your brand, try Howseen.

Which AI crawlers are checked, and where do the names come from? Only official user-agent tokens published by each company: OpenAI (GPTBot, OAI-SearchBot, ChatGPT-User), Anthropic (ClaudeBot, Claude-SearchBot, Claude-User), Perplexity (PerplexityBot, Perplexity-User), Google (Googlebot, Google-Extended), Apple (Applebot-Extended) and Microsoft Bing (Bingbot).

Why does text without JavaScript matter? Vercel's crawler study found that "none of the major AI crawlers currently render JavaScript", with Googlebot and Applebot as exceptions (source). If your content only appears after JavaScript runs, those crawlers see an empty page.

Should I block GPTBot or ClaudeBot? It is your call. These bots collect data for model training; the search and assistant bots (OAI-SearchBot, Claude-SearchBot, PerplexityBot...) are the ones that fetch pages to answer users. That is why training bots weigh 5 points and search bots 25.

What is llms.txt? A proposed standard (llmstxt.org): a Markdown file at the root of your site that sums up who you are and links your key pages for language models. It is optional and not used by every AI, so it weighs 5 points.

Does it use a browser or AI? No. Plain HTTP requests only (no headless browser, no AI model), which keeps each run light and cheap on your Apify usage. It runs with 256 MB of memory.

Is my site's firewall a problem? Some firewalls block automated requests. If a URL returns 403 or a challenge page, the result says so; AI crawlers may be blocked the same way, so check your firewall or CDN bot settings.

---

This checks if AI can read your site. To see if ChatGPT, Gemini, Perplexity and Google AI actually recommend your brand, try Howseen: https://howseen.ai/?utm_source=apify&utm_medium=actor&utm_campaign=ai-visibility-checker