AI coding · Models

Embedding

Also called: Vector embedding, Text embedding, Embeddings

Turn text or code into a numeric vector so semantically similar content sits closer in space, which helps search and dedupe.

In detail

An embedding model does not answer questions. It maps a passage to a list of numbers. Similar meanings land closer together, so you can find related docs or search code by meaning rather than exact keywords.

In AI coding tools, codebase indexing, RAG, and semantic search usually sit on embeddings: chunk files, store vectors, then on a question pull the nearest chunks into context. That role differs from a Large Language Model: embeddings find; the LLM reads and generates.

You rarely compute vectors by hand, but you should know bad chunking or stale indexes make retrieval—and then generation—miss the point.

Developer infoTerm ID, DOM cues, match priority

Term ID: ai-embedding
DOM selectors: No DOM cues. This concept isn't detected directly on a page.
Priority: 1 · when several match at the same level, the higher priority wins
Version: v1 · updated Oct 5, 2026