Kuldeep Paul

AI referrer domains

The hostnames AI assistants send as referrers, so you can separate AI traffic from regular referrals in GA4, Looker Studio or any analytics tool.

Compiled by , data aggregatorUpdated 24 referrer hostsDownload CSV · JSONMethodology

GA4 regex: all AI assistants

^(.*\.)?(chatgpt\.com|chat\.openai\.com|perplexity\.ai|gemini\.google\.com|bard\.google\.com|notebooklm\.google\.com|copilot\.microsoft\.com|copilot\.cloud\.microsoft|edgeservices\.bing\.com|claude\.ai|chat\.deepseek\.com|chat\.mistral\.ai|meta\.ai|grok\.com|you\.com|duck\.ai|poe\.com|phind\.com|chat\.qwen\.ai|kimi\.com|doubao\.com|yiyan\.baidu\.com|huggingface\.co)$

Use with Session source → matches regex in a custom channel group, ordered above Referral. UTM-tagged ChatGPT sessions report chatgpt.com as their source, so the same pattern catches them.

Referrer hosts

ReferrerProductOperatorNotes
chatgpt.comChatGPTOpenAIMain ChatGPT web/app referrer; citation links also carry utm_source=chatgpt.com.
chat.openai.comChatGPT (legacy)OpenAIPre-2024 domain; still appears in historical data.
perplexity.aiPerplexityPerplexity AIIncludes www.perplexity.ai.
gemini.google.comGeminiGoogleSome Gemini clicks arrive as google.com / organic instead.
bard.google.comBard (legacy)GooglePre-February 2024 domain.
notebooklm.google.comNotebookLMGoogle—
copilot.microsoft.comMicrosoft CopilotMicrosoftConsumer Copilot.
copilot.cloud.microsoftMicrosoft 365 CopilotMicrosoftWork Copilot; often stripped by enterprise browsers.
edgeservices.bing.comCopilot in EdgeMicrosoftEdge sidebar Copilot.
claude.aiClaudeAnthropic—
chat.deepseek.comDeepSeekDeepSeek—
chat.mistral.aiLe ChatMistral AI—
meta.aiMeta AIMetaIn-app clicks from WhatsApp/Instagram usually arrive without a referrer.
grok.comGrokxAIGrok inside X arrives as x.com / t.co.
you.comYou.comYou.com—
duck.aiDuck.aiDuckDuckGo—
poe.comPoeQuora—
phind.comPhindPhind—
chat.qwen.aiQwen ChatAlibaba—
kimi.comKimiMoonshot AIAlso kimi.moonshot.cn.
doubao.comDoubaoByteDance—
yiyan.baidu.comERNIE Bot (Yiyan)Baidu—
huggingface.co/chatHuggingChatHugging FaceMatch the path, not just the host.
perplexity.ai/cometCometPerplexity AIComet clicks mostly appear as perplexity.ai or direct.

See which search index and crawlers sit behind each engine.

Setting up an AI channel in GA4

  1. Go to Admin → Data display → Channel groups and copy the default channel group.
  2. Add a new channel named AI assistants with the condition Source matches regex, and paste the pattern above.
  3. Drag the new channel above Referral; GA4 assigns sessions to the first matching channel.
  4. Use the new group in Traffic acquisition reports and explorations. It applies retroactively to existing data.

Google AI Overviews and AI Mode can't be separated this way: their clicks arrive as google.com and are counted as organic search.

About this dataset

What it covers

Hostnames that AI assistants and answer engines send as the referrer when someone clicks a link in an answer, with the product and operator behind each.

How it was built

Compiled from the operators' documentation, analytics vendors' AI-source lists and observed referrers, and cross-checked against the engine reference.

Limitations

Apps and in-app browsers often send no referrer, so some AI traffic always lands in Direct. Referrer policies change; re-check if a source suddenly drops to zero.

Columns in ai-referrers.csv

host
Referrer hostname (or host plus path) to match
product
AI product sending the traffic
operator
Company behind it
note
Caveats for analytics

Questions

How do I track ChatGPT traffic in GA4?

In GA4, create a custom channel group with a channel called "AI assistants" whose condition is Session source matches regex, using the pattern on this page. Place it above Referral so AI sessions aren't counted there. ChatGPT also appends utm_source=chatgpt.com to cited links, so that source appears even when the referrer is stripped.

Why does some AI traffic show up as direct?

Mobile apps, in-app browsers (WhatsApp, Instagram, the ChatGPT and Gemini apps) and some desktop clients don't send a referrer. Those visits land in Direct. UTM tagging like ChatGPT's recovers part of it; the rest can only be estimated.

Does Google AI Overviews or AI Mode traffic have its own referrer?

No. Clicks from AI Overviews and AI Mode arrive as google.com and are reported with ordinary organic search traffic in GA4 and Search Console.

Analysis using this data

  1. The opt-out asymmetry: you can block the small AI engines, not the big oneThe AI surfaces with the most users are the ones publishers can't refuse and can't measure.

How to cite this data

This dataset is free to use, including commercially, under Creative Commons Attribution 4.0. The one rule is attribution:

  1. Credit the source as “Kuldeep Paul, kuldeep.cc” wherever you use a figure, chart or table from it.
  2. Link to this page (https://kuldeep.cc/data/ai-search/referrers), not to the download file, so readers can check the source of every number.
  3. Include the date. These figures change; say “as of Oct 29, 2026” or link to the version you used.
  4. Mark changes. If you filter, recalculate or chart the data yourself, say so (“analysis based on data from…”).
  5. Cite the original source for single figures when you quote one reported number; it is linked in every row.

In running text: according to Kuldeep Paul's AI referrer domains (kuldeep.cc). More in the methodology.

APA

Paul, K. (2026). AI referrer domains [Data set]. kuldeep.cc. https://kuldeep.cc/data/ai-search/referrers

MLA

Paul, Kuldeep. "AI referrer domains." kuldeep.cc, 29 October 2026, https://kuldeep.cc/data/ai-search/referrers.

Chicago

Paul, Kuldeep. 2026. "AI referrer domains." Dataset. kuldeep.cc. Updated Oct 29, 2026. https://kuldeep.cc/data/ai-search/referrers.

BibTeX

@misc{paul2026referrers, author = {Paul, Kuldeep}, title = {AI referrer domains}, year = {2026}, howpublished = {\url{https://kuldeep.cc/data/ai-search/referrers}}, note = {Dataset, CC BY 4.0, updated 2026-10-29} }

HTML link

<a href="https://kuldeep.cc/data/ai-search/referrers">AI referrer domains</a> (data: Kuldeep Paul, kuldeep.cc)