LanceDB

Architecture

Deployment

Security

API Reference

Support

Hybrid Search

Introduction

Effortlessly transform Open Source experience to Enterprise

Quick Start Guide

Performance

Ingest data

Update data

Build an index

Vector Search

Full Text Search

Metadata Filtering

Query Data with SQL

LanceDB APIs let you interact programmatically with your remote database and tables.

List tables, optionally, with pagination.

List Tables

Create a new table.

Will return 400 if table with requested name already exists.


Create a new table

Count the number of rows in a table.

Can optionally pass a predicate to count the rows that match.


Count rows in a table

Rename a table.

Drop a table

If the table does not exist, it will return 200.


Drop a table

Describe a table and return Table Information.

Describe a table

Search for rows, using any combination of vector search, full text
search, hybrid search and SQL filter expressions.


Query

Insert data to a table

Update rows in a table, using SQL expressions and an optional predicate.


Apply a SQL update to a table.

Delete rows from a table using a SQL predicate.

Delete rows from a table

Perform a merge-insert SQL query against a table. This can be used
to perform upserts. To perform an upsert, set `when_matched_update_all`
and `when_not_matched_insert_all` to true.


Merge-insert data to a table

Create an index on a column.

Vector index options are: `IVF_PQ` and `IVF_HNSW_SQ`. The `distance_type`
parameter is required for vector indices.

For [full text search](https://lancedb.github.io/lancedb/fts/), use `FTS`.
For [scalar indices](https://lancedb.github.io/lancedb/guides/scalar_index/),
use `BTREE`, `BITMAP`, or `LABEL_LIST`.


Get Started

Benchmark

Work with data

Hybrid Search