Use the following resources to store and query vector data, decide whether you need an index yet, and find deeper guidance on filtering, tuning, and measuring recall.
- Storing and querying vectors: Add vector columns, insert and query data, decide when to build an index, and maintain it over time.
- Filtering and hybrid search: Combine
ivfpluswithWHERE-clause filters and full-text search. - Tuning and sizing: Choose
listsandprobes, and what to expect for index size and build cost. - Measuring recall and benchmarking: Validate recall before relying on the index in production.