# RAG (Retrieval-Augmented Generation)

> What RAG is, how AI assistants retrieve passages from the web, and why the individual passage counts for AI visibility, not the whole page.

Synonyms: Retrieval-Augmented Generation, RAG, retrieval-based generation

RAG (Retrieval-Augmented Generation) describes how modern AI assistants like ChatGPT, Perplexity or Gemini arrive at their answers. Instead of relying only on their trained knowledge, at the moment of the question they retrieve fitting text passages from an index and build the answer from them — with a source reference. For a business this means: not the whole website is evaluated, but whether individual, self-contained passages answer a concrete question cleanly and machine-readably. Clear structure and self-contained statements decide whether your content gets cited in AI answers. What that splitting into passages looks like in practice is shown in how an AI reads your website.

---
Source: https://btlabs.dev/en/glossary/rag-retrieval-augmented-generation
Last-Modified: 2026-08-10T13:41:38.106Z
Languages: [de](https://btlabs.dev/llms/de/glossary/rag-retrieval-augmented-generation) · [it](https://btlabs.dev/llms/it/glossary/rag-retrieval-augmented-generation)
See also: [llms.txt](https://btlabs.dev/llms.txt) · [ai.txt (Policy)](https://btlabs.dev/ai.txt) · [identity.json](https://btlabs.dev/identity.json)