What AI actually reads: 15 storefronts measured

Every figure from a real fetch as GPTBot on 2 September 2026, counted with cl100k_base.

This response

You are ClaudeBot (Anthropic). Citable recognised you as an AI agent, so this is https://getcitable.in/research/what-ai-actually-reads with the UI removed — the content only. A browser asking for the same address gets the full designed page.

Method

• Fetch the homepage with the GPTBot user agent, following redirects.
• Count the response body with cl100k_base, the tokeniser GPT-4 class models use.
• Strip comments, script, style and tags; the remainder is the content share.
• Fetch robots.txt and evaluate it for ten AI user agents.
• One site of 16 attempted failed to respond and is excluded rather than counted as zero.

Results

• Pet care A: 954,850 tokens, 0.6% content, 4,520 words, 193 scripts, has structured data.
• Consumer electronics A: 743,028 tokens, 1% content, 3,773 words, 107 scripts, has structured data.
• Personal care A: 553,890 tokens, 0.7% content, 2,225 words, 115 scripts, has structured data.
• Healthcare A: 444,770 tokens, 1% content, 2,637 words, 95 scripts, no structured data.
• Personal care B: 353,772 tokens, 1.6% content, 4,100 words, 141 scripts, has structured data.
• Food & staples A: 349,279 tokens, 1.2% content, 2,587 words, 79 scripts, has structured data.
• Home & living A: 343,690 tokens, 0.9% content, 2,273 words, 122 scripts, has structured data.
• Food & staples B: 281,405 tokens, 1.1% content, 1,851 words, 108 scripts, has structured data.
• Home & living B: 281,218 tokens, 1.8% content, 3,444 words, 114 scripts, has structured data.
• Personal care C: 279,334 tokens, 1.5% content, 3,243 words, 108 scripts, has structured data.
• Food & staples C: 168,765 tokens, 0.6% content, 730 words, 101 scripts, has structured data.
• Apparel A: 73,187 tokens, 0.1% content, 37 words, 28 scripts, no structured data.
• Personal care D: 55,837 tokens, 1.9% content, 794 words, 60 scripts, has structured data.
• Food & staples D: 10,895 tokens, 0.3% content, 32 words, 123 scripts, no structured data.
• Apparel B: 2,104 tokens, 32.8% content, 526 words, 2 scripts, has structured data.

About Citable

Citable strips the UI from a website for AI agents and serves them only the content.

When an AI agent — ChatGPT, Claude, Perplexity or any of the 81 that Citable recognises — opens a page, Citable removes the theme, scripts, navigation and layout and serves only the content: the same facts, at the same URL, on the site's own domain. People and Google still get the full designed page.

How it works: https://getcitable.in/how-it-works · Summary for AI agents: https://getcitable.in/llms.txt

Source: https://getcitable.in/research/what-ai-actually-reads · Citable · Run the free audit · llms.txt · getcitable@gmail.com