AI search and GEO glossary

Updated 15 November 2026.

Short, sourced definitions of the terms that come up when you work on how ChatGPT, Claude and Grok search and cite. Each entry answers in one sentence, then links to the guide that goes deeper. The idea most of them connect to is query fan-out.

A

• What is Google AI Mode?

  Google Search's conversational mode that answers harder, multi-part questions with a generated reply and links, running many related searches behind it.

• What are Google AI Overviews?

  Generated summaries that Google Search shows above some results, answering the query in a few lines with links to the pages it drew on.

C

• What is a canonical URL?

  The preferred address for a page when the same or very similar content is reachable at more than one URL, so search engines consolidate signals there.

• What is ChatGPT search?

  The ChatGPT feature that searches the web when a question needs current information, then answers with inline citations and a list of sources.

• What is ChatGPT-User?

  OpenAI's user agent for visits a ChatGPT user sets off, such as opening a page to answer a question, rather than automatic crawling.

• What is Claude-SearchBot?

  Anthropic's crawler for improving the quality of Claude's search results, one of three Anthropic bots that respect robots.txt.

• What is Claude-User?

  Anthropic's user agent for fetching a web page when a person's question to Claude needs that page.

• What is ClaudeBot?

  Anthropic's web crawler that collects public content which may be used to train its models.

• What is a context window?

  All the text a language model can take into account at once when it answers, including the conversation, any fetched pages and its own reply.

• What is crawl budget?

  The set of URLs on a site that a search engine can and wants to crawl in a given time, set by how much load the server can take and how much the engine wants the pages.

D

• What is deep research in ChatGPT?

  A ChatGPT mode that runs many web searches and page reads over several minutes and writes a report with citations, rather than a quick answer.

E

• What is E-E-A-T?

  Short for experience, expertise, authoritativeness and trustworthiness: the qualities Google's search quality raters look for in content, with trust counted as the most important.

• What is an embedding?

  A list of numbers that represents the meaning of a piece of text, so that texts with similar meanings get numbers that sit close together.

F

• What is a featured snippet?

  A Google result that shows a short extract answering the query first, with the page title and link under it, often at the top of the results.

• What is function calling?

  A way for a language model to ask the app it runs in to call a named tool, such as a web search, with arguments it chooses, then use the result in its answer.

G

• What is generative engine optimisation (GEO)?

  Shaping content so AI answer engines find it, use it and cite it.

• What is Google-Extended?

  A robots.txt token that tells Google whether content it crawls may be used to train its Gemini models and to ground answers in Gemini Apps.

• What is Googlebot?

  The name for the web crawlers Google Search uses to fetch pages, in a smartphone and a desktop version that share one robots.txt token.

• What is GPTBot?

  OpenAI's web crawler that collects content which may be used to train its foundation models.

• What is grounding in AI?

  Tying a language model's answer to sources it retrieved and can point to, such as search results, so the answer rests on more than what the model learned in training.

H

• What is an AI hallucination?

  A response from an AI model that presents false or misleading information as if it were fact.

I

• What is indexing?

  The step where a search engine analyses a page it has crawled and stores it, so it can be returned for searches.

• What is IndexNow?

  An open protocol that lets a website notify search engines the moment a page is added, updated or deleted, instead of waiting for a crawl.

J

• What is JSON-LD?

  A JSON-based format for describing things as linked data, used on web pages to tell machines what the page is about in a block of code.

K

• What is a knowledge cutoff?

  The date up to which a language model's training gives it reliable knowledge; anything after it the model can only learn by searching or being told.

L

• What is a large language model (LLM)?

  A neural network trained on huge amounts of text, the model behind ChatGPT, Claude and Grok.

• What is llms.txt?

  A proposed Markdown file at the root of a website, /llms.txt, that gives language models a short summary of the site and links to its most useful pages.

• What is a long-tail keyword?

  A specific, longer search phrase that few people search for on its own but that faces less competition, the opposite of a short, popular head term.

N

• What is noindex?

  A rule, set in a page's robots meta tag or an X-Robots-Tag HTTP header, that tells search engines not to include the page in their index.

O

• What is OAI-SearchBot?

  OpenAI's crawler for ChatGPT search: it lets a site's pages appear in ChatGPT's search answers, and sites that block it are left out.

Q

• What is query expansion?

  Rewriting or adding to a search query, with synonyms, related terms or corrected spellings, so it finds more of the relevant results.

R

• What is retrieval-augmented generation (RAG)?

  A model retrieves documents first, then writes its answer from them.

• What is robots.txt?

  A plain text file at the root of a website that tells crawlers which parts of the site they may and may not visit.

S

• What is a search engine results page (SERP)?

  The page a search engine shows for a query: links, ads, snippets and AI answers.

• What are server-sent events (SSE)?

  A web standard for streaming updates from a server to a page over one connection.

• What is structured data?

  Code added to a page that describes its content in a standard vocabulary, usually schema.org, so machines can tell what the page is and what it contains.

T

• What is a token in AI?

  The unit of text a language model reads and writes, usually a word or part of a word; in English, one token is about four characters.

W

• What is a web search tool in an AI model?

  A tool a language model can call to run web searches during an answer, choosing its own queries and citing the pages it uses.

X

• What is X search in Grok?

  A Grok tool that searches posts on X by keyword, meaning or account, can pull whole threads, and cites the posts it used in the answer.

• What is an XML sitemap?

  A file, usually sitemap.xml, that lists a site's pages and when each last changed, so search engines can find and crawl them.

Missing a term? The blog covers fan-out, citations and GEO in depth.