Can AI answer engines read this page?
Paste a URL. We check its robots.txt against five AI crawler user-agents, then fetch the page the way each of them would. This is an observation, not a score — a page can be allowed and still unreachable, and both states mean something different.
No URL checked yet. The five crawlers below are the ones this tool probes.
| Crawler | Operator | Role | robots.txt | HTTP | Time |
|---|---|---|---|---|---|
| GPTBot | OpenAI | Training agent | Not checked | — | — |
| OAI-SearchBot | OpenAI | Retrieval agent. OpenAI recommends allowing it | Not checked | — | — |
| ChatGPT-User | OpenAI | Live fetch, while a person is asking | Not checked | — | — |
| ClaudeBot | Anthropic | Training agent. Claude-SearchBot is the retrieval one | Not checked | — | — |
| PerplexityBot | Perplexity | Retrieval agent. Perplexity states it is not used to crawl for foundation models | Not checked | — | — |
Tokens and roles read from OpenAI, Anthropic and Perplexity on 8 September 2026. Tokens change; re-read them before you write a rule. Which index a retrieval agent fills is weaker evidence, and which index each assistant searches sets out what is settled and what is not.
What this does not tell you: whether an engine will cite the page. Being fetchable is the first gate, not the outcome.