HL

Open Source // MIT

See what AI sees
when it looks at your page

Paste a URL to see how GPTBot, ClaudeBot, Googlebot, PerplexityBot, and Gemini process your content. No scores — just the ground truth.

01

Robots & Access

See which AI bots your robots.txt allows or blocks — GPTBot, ClaudeBot, PerplexityBot, Googlebot, and more.

02

Per-Bot Extraction

Each bot processes your page differently. See the exact content each one extracts — chunks, Markdown, rendered DOM.

03

Side-by-Side Compare

Compare word counts, truncation, and content visibility across all bots in a single summary view.

Why Does This Matter?

Every AI system processes your page differently. GPTBot doesn't render JavaScript. Googlebot does. ClaudeBot converts HTML to Markdown and truncates at 100KB. ChatGPT reads in 200-word sliding window chunks. Google silently drops everything after 2MB.

A React SPA that ranks #1 on Google may be completely invisible to ChatGPT, Claude, and Perplexity — because those bots see an empty <div id="root"></div>. Headlights makes the invisible visible.

How Each Bot Processes Your Page

ChatGPT

No JS

Plain text, ~200-word sliding window chunks

Context window

Claude

No JS

Turndown HTML → Markdown

100KB text

Google

Full JS

Rendered DOM, passage-level blocks

2MB HTML

Perplexity

No JS

Plain text extraction

Undocumented

Gemini

Full JS (via Google)

Search snippets, passage-level

Snippet-length

Claude Code

No JS

Turndown → Markdown → LLM summary

~100KB pre-summary

The Visibility Gap

Site owners optimize for Google and assume the rest follows. It doesn't. Blocking Google-Extended in robots.txt prevents Gemini from grounding on your page — but has no effect on AI Overviews. Blocking PerplexityBot doesn't stop Perplexity-User from fetching on demand.

The rules are fragmented, underdocumented, and changing fast. Headlights doesn't interpret or score — it shows you the raw truth of what each bot actually ingests.

Quick Reference: Key Constraints

GPTBotNo JS rendering
ClaudeBot100KB Markdown truncation
GooglebotFull JS, 2MB hard cutoff
ChatGPT~200-word sliding windows
GeminiGoogle-Extended controls access
Perplexity-UserIgnores robots.txt

Stop Guessing. Start Seeing.

Enter any URL above to see exactly what each AI bot extracts from your page. No signup. No tracking. Open source.