Weekly report · 2018-W31

The state of AI crawling

Measured on 30,502 reachable domains for the week ending 2018-08-05. Every figure is the result of an automated audit run the same way on every domain: read robots.txt, resolve the rules for 48 tracked agents, then send live requests as four of them.

← 2018-W30  ·  2018-W32 →

30,502
domains measured
6%
block an answer engine
19%
refuse AI agents at the edge
13%
publish an llms.txt

Block rate by crawler with change on the previous week

CrawlerOperatorContent used forBlock rateSitesWeek on week
GPTBot OpenAI training 11.4% 3,465 of 30,503 new
CCBot Common Crawl Foundation archive 11.2% 3,402 of 30,503 new
Bytespider ByteDance training 10.7% 3,273 of 30,503 new
ClaudeBot Anthropic training 10.0% 3,043 of 30,503 new
Google-Extended Google training 9.0% 2,743 of 30,503 new
Meta-ExternalAgent Meta training 9.0% 2,739 of 30,503 new
Amazonbot Amazon training 9.0% 2,734 of 30,503 new
Applebot-Extended Apple training 8.8% 2,670 of 30,503 new
anthropic-ai Anthropic training 4.2% 1,278 of 30,503 new
omgilibot Webz.io training 4.2% 1,271 of 30,503 new
Diffbot Diffbot training 3.9% 1,176 of 30,503 new
omgili Webz.io training 3.8% 1,155 of 30,503 new
cohere-ai Cohere live retrieval 3.8% 1,150 of 30,503 new
SemrushBot-OCOB Semrush training 3.7% 1,135 of 30,503 new
PerplexityBot Perplexity answer index 3.6% 1,107 of 30,503 new
ChatGPT-User OpenAI live retrieval 3.6% 1,100 of 30,503 new
FacebookBot Meta training 3.3% 992 of 30,503 new
YouBot You.com answer index 2.8% 868 of 30,503 new
ImagesiftBot ImageSift (Hive) training 2.8% 839 of 30,503 new
Timpibot Timpi training 2.7% 835 of 30,503 new
Scrapy Zyte (open-source framework) training 2.3% 710 of 30,503 new
OAI-SearchBot OpenAI answer index 2.3% 688 of 30,503 new
AI2Bot Allen Institute for AI training 2.2% 686 of 30,503 new
cohere-training-data-crawler Cohere training 2.0% 597 of 30,503 new
DuckAssistBot DuckDuckGo live retrieval 1.7% 530 of 30,503 new
Meta-ExternalFetcher Meta live retrieval 1.7% 529 of 30,503 new
PanguBot Huawei training 1.7% 519 of 30,503 new
Claude-SearchBot Anthropic answer index 1.6% 482 of 30,503 new
Claude-User Anthropic live retrieval 1.6% 475 of 30,503 new
Kangaroo Bot Kangaroo LLM training 1.5% 460 of 30,503 new
Perplexity-User Perplexity live retrieval 1.5% 448 of 30,503 new
MistralAI-User Mistral AI live retrieval 1.4% 428 of 30,503 new
Applebot Apple answer index 1.2% 355 of 30,503 new
Google-CloudVertexBot Google live retrieval 1.2% 353 of 30,503 new
Meta-WebIndexer Meta answer index 1.1% 349 of 30,503 new
GoogleOther Google training 1.1% 333 of 30,503 new
TikTokSpider ByteDance live retrieval 0.8% 255 of 30,503 new
Amzn-SearchBot Amazon answer index 0.7% 220 of 30,503 new
Amzn-User Amazon live retrieval 0.4% 130 of 30,503 new
MistralAI-Index Mistral AI answer index 0.2% 69 of 30,503 new
facebookexternalhit Meta live retrieval 0.2% 57 of 30,503 new
MistralAI-Training Mistral AI training 0.2% 56 of 30,503 new
Bingbot Microsoft answer index 0.1% 41 of 30,503 new
Googlebot-News Google answer index 0.1% 38 of 30,503 new
msnbot Microsoft answer index 0.1% 26 of 30,503 new
Meta-ExternalAds Meta training 0.1% 20 of 30,503 new
Googlebot Google answer index 0.1% 16 of 30,503 new
OAI-AdsBot OpenAI live retrieval 0.0% 3 of 30,503 new

Trend share of measured sites

% of sites20151050% of sites08-2208-3109-09block an answer engineblock a training crawlerpublish llms.txt

Where the blocking is share blocking an answer engine, by top-level domain

bar chart.uk (n=452)17%.fr (n=325)16.9%.de (n=838)16.5%.jp (n=421)9%.com (n=14987)6.6%.br (n=379)6.1%.org (n=1682)5.2%.net (n=1324)4.1%.ru (n=1166)3%.in (n=305)2.6%.edu (n=478)2.3%.io (n=657)1.8%

Only TLDs with at least 40 measured domains are shown, and each bar carries its sample size. A TLD is a weak proxy for jurisdiction. Full segmentation.

Changes recorded this week 0 total

No crawler policy changes were recorded in this week.

Full change log · RSS

Cite this CC BY 4.0

The figures on this page are frozen to the week shown and will not be edited after publication. Corrections are appended, not overwritten.

Crawl Census, "The state of AI crawling, 2018-W31", 2018-08-05. https://crawlcensus.com/report/2018-W31

Machine-readable: census totals, full corpus, crawler registry. Method: methodology.