![[str_idx_ivfflat.png|256]]
- **Description**: Divides vector space into clusters (lists) and searches only
the most relevant ones. Faster to build and uses less memory than HNSW, but
with lower query throughput.
- **Extension**: Requires `pgvector`.
- **Supported Operators**: `<->`, `<#>`, `<=>`, `<+>` (L1 distance).
- **Key Parameters**:
- `lists`: Number of clusters to create (training required).
- `probes`: Number of clusters to search during a query (runtime tuning).
- **Metaphor**: A city map divided into distinct neighborhoods. Instead of
searching every house in the city, you identify the 2 or 3 neighborhoods most
likely to contain your target and only search there.