> ## Documentation Index
> Fetch the complete documentation index at: https://docs.anymorph.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# About AI Engines

> Engine-specific AI search characteristics for GEO measurement and strategy

AI engines can answer the same prompt with different retrieval systems, citation interfaces, query expansion patterns, and personalization signals. GEO performance therefore should not be interpreted as a single universal score. It should be broken down by engine: which source pool the answer came from, which page types were cited, and how the brand was recommended or omitted.

This guide summarizes the product background, usage scale, citation/source behavior, query fan-out patterns, commonly cited source types, and measurement priorities for major AI engines.

## Engine pages

| Engine / surface   | Core question                                                           | Page                                                          |
| ------------------ | ----------------------------------------------------------------------- | ------------------------------------------------------------- |
| Google AI Overview | Which pages survive as sources inside Google's SERP AI summary?         | [Google AI Overview](/concepts/ai-engines/google-ai-overview) |
| Google AI Mode     | Which intent branches are missed by Gemini-style query fan-out?         | [Google AI Mode](/concepts/ai-engines/google-ai-mode)         |
| Gemini             | Is the brand correctly understood and recommended even without sources? | [Gemini](/concepts/ai-engines/gemini)                         |
| ChatGPT            | Did the answer use Bing-backed retrieval or parametric knowledge?       | [ChatGPT](/concepts/ai-engines/chatgpt)                       |
| Perplexity         | Which sources and passages survive into inline citations?               | [Perplexity](/concepts/ai-engines/perplexity)                 |
| Claude             | Which claims and sources are treated as high-confidence evidence?       | [Claude](/concepts/ai-engines/claude)                         |
| Grok               | How is the brand described in X and real-time social context?           | [Grok](/concepts/ai-engines/grok)                             |
| Microsoft Copilot  | How should Bing visibility and M365/Graph visibility be separated?      | [Microsoft Copilot](/concepts/ai-engines/microsoft-copilot)   |

## Comparison dimensions

| Dimension                | Why it matters                                                                                                                       |
| ------------------------ | ------------------------------------------------------------------------------------------------------------------------------------ |
| Product surface          | SERP summaries, assistants, and enterprise copilots expose sources and shape user behavior differently.                              |
| Usage scale              | MAU, WAU, DAU, and query volume indicate reach. When official numbers are unavailable, distribution base and usage context matter.   |
| Retrieval substrate      | Google, Bing, proprietary indexes, X, and Microsoft Graph create different citation candidate pools.                                 |
| Citation/source behavior | Citation-first engines and assistants where sources may appear or be omitted require different KPIs.                                 |
| Query fan-out            | Brands can disappear in comparison, alternative, pricing, review, or freshness branches even when the canonical query looks healthy. |
| Source type preference   | Wikipedia, Reddit, YouTube, review platforms, official docs, and owned answer surfaces carry different weight by engine.             |