Kage — context augmented generation — is where you load documents into the prompt.
It uses more of the context window, so it’s fast but has limits.
RAG — retrieval augmented generation — searches live.
It grabs the most relevant info from a vector database.
Smaller prompt, more compute and retrieval.
Kage gives speed. RAG brings scale.
Together, you get context and intelligence.
