RadarTrek
Work with us
Home/Glossary/RAG
AI & LLMs

RAG

RAG

Retrieval-Augmented Generation — giving an LLM your own documents to read before it answers.

Reviewed by the RadarTrek editorial team · June 2026

RAG (Retrieval-Augmented Generation) retrieves the most relevant chunks of your own content via similarity search, then includes them in the prompt so the model answers grounded in your actual data instead of only its training data. It's the standard alternative to fine-tuning when the real problem is "the model doesn't know my content," not "the model has the wrong writing style."

RAG answers from YOUR documents, not just training data User asks a question Embed query → search vector database finds similar chunks Retrieved chunks doc excerpt 1 doc excerpt 2 doc excerpt 3 LLM reads question + chunks, generates answer grounded in your docs Question Retrieve Augment Generate

Why it matters

  • —RAG is almost always the right choice over fine-tuning when your data changes frequently or you need citations.
  • —RAG reduces but does not eliminate hallucination — if retrieval misses the right chunk, the model may still guess.
  • —A RAG pipeline has three phases: ingest (chunk + embed + store), retrieve (embed the query, find similar chunks), and generate.

Where to learn this

🎓

What RAG Is — and Why Fine-Tuning Is Usually the Wrong Answer

RAG and Vector Search course

This is the exact lesson that covers this term in depth — with examples, diagrams, and a hands-on exercise.

Related terms

EmbeddingVector DatabaseFine-TuningHallucination
← All glossary terms

RadarTrek Intel — monthly score updates

We track 40+ tools so you don't have to. Score changes, new tools, and new guides — once a month, no spam.