Database in Python

6 projects in Database & Storage. Most starred: duckdb, chromadb and zvec.

These databases run inside your Python process, with no server to deploy. DuckDB is an analytical SQL engine: it queries Parquet, CSV and JSON files directly, as well as pandas and Polars DataFrames in memory, and is a common replacement for heavy DataFrame code on large local datasets. chDB embeds the ClickHouse engine in the same way, which helps if you already write ClickHouse SQL.

For embeddings and retrieval, LanceDB stores vectors and their data in the columnar Lance format on local disk or object storage, Chroma offers a simple API aimed at retrieval-augmented generation, and zvec is a lightweight in-process vector store. TinyDB keeps JSON documents in a single file and is pure Python, which suits small tools and prototypes but not concurrent writes or large data.

Analytical2

duckdbC++ core

An in-process SQL OLAP database management system; optimized for analytics and fast queries, similar to SQLite but for analytical workloads.

42.5kupdated 1d ago2.0.0

chdb

In-process OLAP SQL engine with the full ClickHouse dialect, zero-copy pandas/Arrow interop, and federation to remote ClickHouse clusters via remoteSecure().

2.9kupdated 11d ago4.4.0

Vector3

lancedbRust core

A developer-friendly embedded retrieval database for multimodal AI.

11.8kupdated today0.40.0

chromadbRust core

An open-source embedding database for building AI applications with semantic search.

29.6kupdated 1d ago1.5.9

zvecC++ core

A lightweight, in-process vector database that embeds directly into applications.

16.1kupdated 2d ago0.8.0

Key-Value & Document1

tinydb

A tiny, document-oriented database.

7.6kupdated 9d ago4.9.0
Esc

Type to search every entry. ↑↓ to move, Enter to open.