What the ai bot access checker checks

18 crawlers. One verdict each.

Every AI crawler that matters, evaluated against the exact URL you enter, with the rule that decided it.

18 crawlers3 roles
  • OAI-SearchBotOpenAIsearch
  • Claude-UserAnthropicbrowse
  • PerplexityBotPerplexitysearch
  • Google-ExtendedGoogletrain

18 AI crawlers

0 pts

OpenAI, Anthropic, Perplexity, Google, Microsoft, Apple, Amazon, Meta, DuckDuckGo, Mistral, ByteDance and Common Crawl, grouped into AI search, live browsing and model training.

User-agent: *
Disallow: /private/
Allow: /private/helplongest match wins
User-agent: PerplexityBot
Disallow: /blocks everything

Real robots.txt evaluation

0 pts

Parses every group and applies the standard: the most specific user-agent group wins, the longest matching path wins, wildcards and $ anchors are honoured.

/robots.txt18 bots
  • User-agent: OAI-SearchBotAllow /
  • User-agent: ClaudeBotAllow /
  • User-agent: PerplexityBotAllow /
  • User-agent: GPTBotDisallow /

Verdict for the exact URL

0 pts

Allowed, restricted (blocked on this path but not site-wide) or blocked, with the deciding rule printed next to every crawler.

<html lang="en">
<link rel="canonical">
<meta name="description">
<main>
text 18%

Page-level directives

0 pts

Reads the robots meta tag and the X-Robots-Tag header for noindex, noai and noimageai, which override robots.txt for the page itself.

output
CopyDownload
# Acme
> Acme makes widgets for teams.
## Docs
- [Getting started](/docs/start): Install in 5 min
- [API](/docs/api): Endpoints and auth

Ready-to-paste fix

0 pts

Generates the robots.txt lines that reopen the door for every AI search or browsing crawler you are currently blocking.

/sitemap.xml1,204 URLs
<urlset>
<loc>
200
<loc>
404
<lastmod>2026-09-01

Sitemap declarations

0 pts

Lists the Sitemap: lines in robots.txt so you can confirm crawlers are pointed at your URL inventory.

Why it matters

Blocking the wrong bot is silent. Until the traffic is gone.

Most sites that block AI crawlers did it once, in 2023, with a rule that now removes them from ChatGPT search, Perplexity and Claude. The crawlers have changed. The rules have not.

01

Search crawlers are not training crawlers

OpenAI uses GPTBot for training but OAI-SearchBot for ChatGPT search. Anthropic splits ClaudeBot from Claude-SearchBot. Blocking the training bot is fine. Blocking the search bot makes you invisible in answers.

AI searchModel training
02

One wildcard rule can block everyone

A Disallow under User-agent: * applies to every crawler without its own group, including bots launched after you wrote the file. The checker shows which rule actually applies to each bot.

Real robots.txt evaluation
03

The page can override the file

Even with a permissive robots.txt, a noindex meta tag or an X-Robots-Tag header keeps a page out of the index that AI answers draw from. Both are checked for the URL you enter.

Page-level directives
How to improve

Open the door in order. Search crawlers first, training crawlers on your terms.

  1. 1

    Allow the AI search crawlers explicitly

    Add groups for OAI-SearchBot, ChatGPT-User, Claude-SearchBot, Claude-User, PerplexityBot and Perplexity-User with Allow: /. Explicit groups win over the wildcard group.

  2. 2

    Decide on training crawlers separately

    GPTBot, ClaudeBot, Google-Extended, CCBot, Applebot-Extended, meta-externalagent and Bytespider only collect training data. Block them if you wish; it does not affect AI search visibility.

  3. 3

    Keep Googlebot and Bingbot open

    Google AI Overviews and Bing Copilot use the classic search crawlers. A block here removes you from both traditional and AI results.

  4. 4

    Check page-level directives

    Remove noindex from pages you want cited, and drop noai unless you deliberately opt out of AI answers.

  5. 5

    Re-test after every deploy

    robots.txt is easy to overwrite from a staging config. Paste the URL again after each release; results refresh within a few hours.

Score after fixes
41 → 93
F
D
C
B
A
robots.txt
llms.txt
FAQPage

Open the door, then give them something to cite

SiteLift publishes AI-ready articles with structured data and FAQs on your site, and distributes them so the crawlers you just unblocked have a reason to come back.

No credit card required

Frequently Asked Questions

Everything you need to know about the ai bot access checker.

OAI-SearchBot, ChatGPT-User and GPTBot (OpenAI); Claude-SearchBot, Claude-User and ClaudeBot (Anthropic); PerplexityBot and Perplexity-User; Googlebot and Google-Extended; Bingbot; DuckAssistBot; Applebot-Extended; Amazonbot; meta-externalagent; CCBot; Bytespider; and MistralAI-User. Eighteen in total, grouped into AI search, live browsing and model training.