Semantic (vector) search
infino_semantic_searchSearch by meaning to retrieve paraphrases and synonyms, ranked by vector similarity; optionally pre-filter rows by a keyword condition first.
Instructions
Use when searching for a concept by meaning and the exact wording is unknown — this retrieves paraphrases and synonyms, not just literal matches. Embeds the query with a local model (no API key) and ranks a table's embedding column by vector similarity. Each hit carries a score that is a DISTANCE (lower is closer) plus the columns you project ('columns'; the full text column by default). Optional 'filter' restricts the ranking to rows whose keyword column matches a predicate first (a pushdown pre-filter, e.g. semantic search only within rows tagged 'billing'). For exact terms use infino_keyword_search; when the query has both literal terms and an intent use infino_hybrid_search.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| k | No | Maximum results. | |
| query | Yes | Query text; embedded and matched by vector similarity. | |
| table | Yes | Table to search. | |
| column | No | Text column to return with each hit; inferred if omitted. | |
| filter | No | Pre-filter: rank the kNN only among rows whose FTS 'column' matches 'query' (a pushdown pre-filter, not a post-filter on the results). | |
| columns | No | Which of the table's columns each hit returns, with full values (a projection passed straight to the engine). Defaults to the text column; '_id' and 'score' are always included. Any column works: ['id'] for compact hits at a large k, ['id', 'text'] to get the full text alongside an id to cite, ['title', 'created_at'] for metadata. Nothing is truncated; read fewer columns or a smaller k to keep results small. | |
| vectorColumn | No | Vector column to search; inferred if omitted. |