Weekly report · 2015-W28

The state of AI crawling

Measured on 30,350 reachable domains for the week ending 2015-07-12. 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.

← 2015-W27  ·  2015-W29 →

30,350
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 9.3% 2,833 of 30,345 new
CCBot Common Crawl Foundation archive 9.1% 2,773 of 30,345 new
Bytespider ByteDance training 8.7% 2,638 of 30,345 new
ClaudeBot Anthropic training 7.9% 2,406 of 30,345 new
Meta-ExternalAgent Meta training 7.0% 2,130 of 30,345 new
Amazonbot Amazon training 7.0% 2,115 of 30,345 new
Google-Extended Google training 6.9% 2,098 of 30,345 new
Applebot-Extended Apple training 6.8% 2,061 of 30,345 new
anthropic-ai Anthropic training 4.1% 1,257 of 30,345 new
omgilibot Webz.io training 4.1% 1,233 of 30,345 new
cohere-ai Cohere live retrieval 3.8% 1,168 of 30,345 new
Diffbot Diffbot training 3.8% 1,143 of 30,345 new
SemrushBot-OCOB Semrush training 3.7% 1,130 of 30,345 new
omgili Webz.io training 3.7% 1,119 of 30,345 new
PerplexityBot Perplexity answer index 3.5% 1,067 of 30,345 new
ChatGPT-User OpenAI live retrieval 3.5% 1,060 of 30,345 new
FacebookBot Meta training 3.2% 957 of 30,345 new
YouBot You.com answer index 2.9% 878 of 30,345 new
ImagesiftBot ImageSift (Hive) training 2.7% 827 of 30,345 new
Timpibot Timpi training 2.7% 816 of 30,345 new
Scrapy Zyte (open-source framework) training 2.4% 725 of 30,345 new
AI2Bot Allen Institute for AI training 2.2% 669 of 30,345 new
OAI-SearchBot OpenAI answer index 2.2% 666 of 30,345 new
cohere-training-data-crawler Cohere training 2.0% 599 of 30,345 new
Meta-ExternalFetcher Meta live retrieval 1.8% 536 of 30,345 new
DuckAssistBot DuckDuckGo live retrieval 1.7% 512 of 30,345 new
PanguBot Huawei training 1.6% 498 of 30,345 new
Claude-SearchBot Anthropic answer index 1.6% 471 of 30,345 new
Claude-User Anthropic live retrieval 1.5% 470 of 30,345 new
Kangaroo Bot Kangaroo LLM training 1.5% 448 of 30,345 new
Perplexity-User Perplexity live retrieval 1.4% 440 of 30,345 new
MistralAI-User Mistral AI live retrieval 1.4% 422 of 30,345 new
Google-CloudVertexBot Google live retrieval 1.2% 361 of 30,345 new
Meta-WebIndexer Meta answer index 1.2% 353 of 30,345 new
GoogleOther Google training 1.1% 339 of 30,345 new
Applebot Apple answer index 1.1% 333 of 30,345 new
TikTokSpider ByteDance live retrieval 0.8% 254 of 30,345 new
Amzn-SearchBot Amazon answer index 0.7% 211 of 30,345 new
Amzn-User Amazon live retrieval 0.5% 139 of 30,345 new
MistralAI-Index Mistral AI answer index 0.2% 73 of 30,345 new
MistralAI-Training Mistral AI training 0.2% 67 of 30,345 new
facebookexternalhit Meta live retrieval 0.2% 58 of 30,345 new
Bingbot Microsoft answer index 0.1% 39 of 30,345 new
Googlebot-News Google answer index 0.1% 38 of 30,345 new
msnbot Microsoft answer index 0.1% 26 of 30,345 new
Meta-ExternalAds Meta training 0.1% 19 of 30,345 new
Googlebot Google answer index 0.1% 16 of 30,345 new
OAI-AdsBot OpenAI live retrieval 0.0% 5 of 30,345 new

Trend share of measured sites

% of sites20151050% of sites08-2209-0409-17block 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=447)17.4%.fr (n=328)16.5%.de (n=836)16.3%.jp (n=412)9%.com (n=14867)6.5%.br (n=382)6%.org (n=1679)5.4%.net (n=1312)3.7%.ru (n=1165)3%.in (n=301)2.7%.edu (n=475)2.3%.io (n=659)2%

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, 2015-W28", 2015-07-12. https://crawlcensus.com/report/2015-W28

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