Semantic search
Semantic search finds results that match a query's meaning instead of its exact words, using embeddings to compare meaning directly.
Category
Industry concept
See all terms
What is semantic search?
Semantic search finds results based on what a query means rather than which exact words it contains. It works by turning both the query and the searchable content into embeddings and returning whatever sits closest to the query in that meaning space.
This solves a real limit of exact-match search: a person rarely phrases a question the same way a document phrases its answer, so keyword search alone misses correct results that use different words for the same thing.
How does semantic search differ from hybrid search?
Semantic search relies entirely on meaning, so it can miss an exact term, a product code, a name, or an acronym, that carries little meaning of its own but still needs to match precisely. Hybrid search runs semantic and keyword search together and combines the results, so an exact match on a rare term and a conceptual match on a rephrased question both have a path to surface.
How does semantic search work?
A semantic search system embeds every document in advance and stores those vectors in an index. At query time it embeds the incoming query the same way, compares that vector against the stored ones by a distance measure, and returns the closest matches.
Qontext's context repository exposes hybrid search as one of its core read operations, alongside operations like grep and find, so an agent gets both the exact identifier and the conceptual match in one call instead of running two separate searches and merging them by hand.