---
title: "What Is an LLM Citation? Definition & How to Earn One"
description: "An LLM citation is when an AI model attributes or links your content as a source in its answer. Distinct from being merely retrieved — and the real goal of AI optimization."
canonical: https://aiovsseo.com/glossary/llm-citation.html
date: 2026-06-07
---
# What is LLM citation?

Definition

An LLM citation is when an AI model names, links or attributes your content as a source in the answer it gives a user. It is distinct from being retrieved — models use far more sources than they visibly cite.

An **LLM citation** is the visible attribution an AI model gives a source when it answers — a link or named reference the user can follow. It is the AI-era equivalent of an impression-plus-credit, and the central goal of [AI Optimization](/glossary/ai-optimization-aio.html).

## Retrieved is not cited

Models read many more pages than they credit. By some analyses ChatGPT visibly cites only around 15% of the URLs it retrieves; the rest inform the answer silently. Being findable is table stakes; being *quotable* converts retrieval into a citation.

## How to earn citations

Lead with a direct answer, use clear self-contained sections, provide specific dated statistics, and build entity authority. See [what is GEO](/articles/what-is-geo.html) for the scoring model.

## Frequently asked questions

**What is the difference between being retrieved and being cited?**

Retrieval means the model fetched and read your page; citation means it visibly credited you in the answer. Models cite only a fraction of what they retrieve, so quotability matters as much as findability.

**How can I track LLM citations of my brand?**

Run your top category prompts across ChatGPT, Claude, Perplexity and Google AI surfaces and log which sources each names. That transcript is your AI citation baseline.
