Measured on our own site, 2026-10-03

We recorded 44,648 Meta-labeled requests.
No Meta referrals were recorded in this snapshot.

Over about 20 hours, requests labeled as Meta’s training crawler reached 1,576 pages of limitsregistry.com — roughly 28 times each. Crawler request counts and human referral visits measure different things. This dated snapshot shows why we inspect both.

limitsregistry.com is ours. We are not borrowing a customer’s traffic to make a point.

The measurement

Every AI crawler that has read one small website

Retained requests through October 3, 2026, grouped by detected crawler identity. A crawler name is a classification, not independent proof of the operator.

CrawlerOperatorWhat it doesRequests
Meta-ExternalAgentMetaUsed for training; not a referral channel.44,648
OAI-SearchBotOpenAIIndexes for an AI search product. Can send visitors.23,478
GPTBotOpenAIUsed for training; not a referral channel.17,160
AmazonbotAmazonIndexes for an AI search product. Can send visitors.6,576
PerplexityBotPerplexityIndexes for an AI search product. Can send visitors.3,240
ChatGPT-UserOpenAIFetches a page for someone asking a question. Can send visitors.686
ApplebotAppleIndexes for an AI search product. Can send visitors.221
DuckAssistBotDuckDuckGoFetches a page for someone asking a question. Can send visitors.44
Claude-UserAnthropicFetches a page for someone asking a question. Can send visitors.8
ClaudeBotAnthropicUsed for training; not a referral channel.1
Total AI crawler requests96,062

96,062

AI crawler requests received

3

visits sent back by an AI product, all time — all from chatgpt.com

0

recorded Meta referrals through the measurement date

Owner-operated installation

How we measure Limits Registry

A real site we own, not a customer testimonial or an income forecast.

  1. Route the owned domain through the crawler edge integration to record server-side requests. A browser script alone cannot observe crawlers that do not execute JavaScript.
  2. Install the browser tracker separately to record page journeys and referral visits. Compare these observations with crawler events rather than treating requests as people.
  3. For the payment pilot, offer enrollment, issue an access credential, and meter successful authenticated page requests. A Stripe-confirmed collection is separate from accrued usage.

Snapshot measured 2026-10-03: 96,062 crawler requests and 3 recorded AI referrals. Labels can be spoofed; retention, blocked scripts and missing referrers limit coverage. These figures do not establish independent buyers or collected revenue.

Cite this example: VisitorPing, “Limits Registry crawler observation,” measured October 3, 2026. Permanent measurement link.

What this does and does not show

The honest reading

A single site, published whole, with the parts that argue against us left in.

This is one small website. Over the same period it received 29 visits from Google. Do not read 96,062 requests as evidence of scale — read it as the ratio between what was taken and what came back.

Not every crawler is the same. OAI-SearchBot, Applebot and Amazonbot index for search products that genuinely can return readers. Blocking those costs you something. Training requests are not reader visits. Missing referrer data also means recorded referrals are not a complete measure of influence.

Paid access needs an enrolled operator. The Limits Registry pilot now uses VisitorPing credentials and Stripe billing. Successful authenticated page requests accrue usage; blocked requests are not purchases. This October 3 traffic snapshot predates the payment rollout and is not evidence of revenue. Customer payouts and general availability for other websites are not offered here.

We found this the hard way. The crawl pushed our own database usage from roughly 1,700 queries an hour to 29,000 and triggered a warning from our host. The feature exists because the surprise was expensive.

What you can do about it

See them, choose, and set your terms

See every crawler by name

Meta, OpenAI, Perplexity, Amazon, Apple, Anthropic and the rest, separated into the ones that train models and the ones that can return a reader. Most analytics filter all of this out as bot traffic before you ever see it.

Decide who may read what

Allow a crawler, block it, or leave it alone, per crawler and per site. robots.txt, your licence and your sitemap always stay readable — a crawler that cannot read your terms cannot honour them.

Put a price on it

Our Limits Registry pilot offers enrolled operators paid access through VisitorPing and Stripe: $0.01 per successful authenticated page request. Unenrolled crawlers subject to the payment gate receive a payment-required response, not a charge.

Find out who is reading your website.

Add your site, install the tracker, and see the crawlers by name.

No credit card requiredFigures measured 2026-10-03 on limitsregistry.comCheck what your robots.txt tells AI crawlers, free