The vocabulary around AI search shifts fast and half the terms overlap. This page pins down the ones that matter — what each term means, how it differs from its neighbors, and where to read the long version. Every entry has a stable anchor you can link to.
Each definition is written to be quotable on its own. The related link goes to the deep-dive.
Generative Engine Optimization is the practice of optimizing content so that generative AI systems — ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews — mention, cite, or recommend it in their answers. Where SEO targets a ranked list of links, GEO targets inclusion in a single synthesized answer. Progress is measured with mention rate, answer rank, and citation frequency per AI surface.
Related: Generative engine optimization: the complete 2026 guide
Answer Engine Optimization is optimizing content to be selected as the direct answer by answer engines: AI chatbots, voice assistants, and featured-snippet-style results. It emphasizes atomic, quotable passages, question-shaped headings, and structured data that engines can lift verbatim. AEO overlaps heavily with GEO — AEO focuses on the answer format, GEO on the generative engines producing it.
Related: What is answer engine optimization (AEO)? A beginner's guide
LLMO is the umbrella term for increasing a brand's presence in large language model outputs, covering both what a model learned during pretraining and what it retrieves at query time via web search or RAG. Because pretraining influence takes months to land, most LLMO work in practice targets the retrieval path: being fetched and cited when the model searches. LLMO, GEO, and LLM SEO are used near-interchangeably.
Related: How to build an LLMO tool
AI Overviews are Google's AI-generated summary block shown above the organic results on the standard SERP. Each overview synthesizes an answer from multiple sources and links the pages it drew from, so a citation slot there sits above position 1. They render on a subset of queries — heavily on informational ones — and the cited set changes as the underlying Gemini models and retrieval change.
Related: Google AI Overviews explained: what they are and why they matter
AI Mode is Google's dedicated chat-style search surface — a separate conversational tab, distinct from the AI Overviews block on the classic SERP. It runs query fan-out under the hood and returns a synthesized markdown answer with inline citations, tables, and shopping cards. There is no official API for it; tracking AI Mode means capturing the rendered surface itself.
Related: How MentionsAPI measures AI Mode and other live surfaces
Query fan-out is the technique where an AI search system expands one user question into multiple background sub-queries, retrieves results for each, and synthesizes a single answer from the combined set. ChatGPT search, Google AI Mode, and Perplexity all do this. It means a page can be cited for queries the user never typed, so visibility tracking has to account for the sub-queries, not just the head term.
Related: We checked: ChatGPT API misses 96% of what real users see
Share of voice in AI answers is the percentage of AI-generated responses, across a tracked query set, that mention or cite a given brand relative to its competitors. It is the AI-search analogue of SERP share of voice, computed from mention rate and rank position per surface. Because LLM outputs are non-deterministic, a meaningful share-of-voice number requires sampling each query multiple times rather than trusting a single run.
Citation tracking is monitoring which URLs AI engines cite as sources when generating answers. Every engine exposes citations differently — inline links, footnotes, reference lists — and the cited set often differs between an engine's API and its user-facing UI. Tracking normalizes those into structured records (URL, rank, surrounding context) so you can tell which pages actually earn placement in AI answers.
Related: AI Citation Tracking API
AI visibility is a measure of how often and how prominently a brand appears in AI-generated answers across surfaces like ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, and Bing Copilot. It typically decomposes into mention rate, answer rank, sentiment, and citation count per surface. It is the metric GEO and AEO work is trying to move.
Related: What is AI visibility? How to measure it across LLMs
llms.txt is a proposed plain-markdown file served at a site's root that gives LLMs a curated map of the site's most important content. Modeled loosely on robots.txt, it lists key pages with short descriptions so AI crawlers and agents can find canonical content without parsing full HTML. Adoption across major AI crawlers is still uneven, but the file costs almost nothing to serve.
Related: What is llms.txt? How to create one for AI crawlers
One /v1/check call returns mention, rank, and citations for your brand across ChatGPT, Claude, Gemini, Perplexity, AI Overviews, AI Mode, and Bing Copilot. 300 free signup credits, no card.