Database packages

Kuzu is an embedded graph database that executes Cypher queries on property graphs, supporting full-text search and vector indices without requiring a separate server process.

However, be aware that the maintainers are no longer actively developing new features—use it for stable, well-tested workloads, not for cutting-edge functionality or…

Provides a DB-API 2.0 compatible Python interface to SingleStore, a distributed SQL database, supporting both MySQL protocol and HTTP connections with optional C acceleration for faster data reading.

Install it if you need a Python driver for SingleStore or are building vector search or real-time analytics applications on that platform.

Provides a MongoDB-backed checkpoint storage implementation for LangGraph, enabling state persistence and recovery in agentic workflows.

Astroquery provides a Python interface to query and retrieve data from online astronomical databases and web services, with sub-packages for each service like SIMBAD, VizieR, and others.

Install it if you need to query astronomical databases from Python; skip it only if you work exclusively with local data or have no need for online lookups.

RMM provides a common interface for customizing GPU and host memory allocation, with implementations for device memory pooling, pinned host memory, and other allocation strategies.

cuDF is a GPU-accelerated DataFrame library that provides pandas-like APIs for tabular data processing on NVIDIA GPUs, with optional zero-code-change acceleration for existing pandas code via cudf.pandas.

Install only if you have an NVIDIA GPU with compatible CUDA 12 drivers and can manage the 18 runtime dependencies (including CUDA toolkit and GPU libraries).

libkvikio-cu12 provides Python and C++ bindings to NVIDIA's cuFile API, enabling high-performance GPU-accelerated file I/O with support for GPUDirect Storage and seamless host/device memory handling.

pylibcudf-cu12 provides Python bindings for libcudf, a CUDA C++ library that accelerates tabular data operations on NVIDIA GPUs using Apache Arrow data structures.

Provides Active Record-style mixins for SQLAlchemy ORM, enabling Django-like query syntax, eager loading, and CRUD operations on model classes.

MariaDB Connector/Python provides a DB API 2.0-compliant interface for Python programs to connect to and query MariaDB and MySQL databases using a C-based client library.

A Singer target that loads data into Snowflake from any Singer tap, supporting append, upsert, and overwrite load methods with schema flattening and stream mapping.

Install it if you are using Singer taps and need a reliable Snowflake destination; it is the standard choice for this integration pattern.

Connects PyCasbin access control policies to SQLAlchemy-supported databases, allowing policies to be loaded from and persisted to PostgreSQL, MySQL, SQLite, Oracle, SQL Server, Firebird, or Sybase.

However, note that maintenance is dormant (no release in 767 days), so if you encounter bugs or need features, you may need to fork or patch it yourself.

Provides detailed Unicode character properties and metadata from the Unicode Character Database with human-readable aliases, as an alternative to Python's standard library unicodedata module.

No, not recommended for new projects.

Provides base classes, decorators, and utilities for building Apache Superset backend extensions, including REST API endpoints, background tasks, and MCP tools.

librmm-cu12 provides GPU memory allocation and management for CUDA 12 workloads, offering customizable device and host memory pools to optimize GPU-centric applications.

Manages Elasticsearch indices and snapshots through automated actions like deletion, optimization, and retention policies based on age, size, or other criteria.

Install it if you manage Elasticsearch indices and need automated lifecycle policies.

Adds SQL magic commands to IPython and Jupyter notebooks, letting you query databases directly from cells using SQLAlchemy connection strings.

However, the dormant maintenance status (last release over a year ago, Alpha classification) means you should verify compatibility with your Python and IPython…

dmPython is a native Python driver for connecting to Dameng databases, implementing the Python DB API 2.0 specification to enable direct database access from Python applications.

Install it if you need to connect Python applications to Dameng databases.

Acsylla is an async Python client library for Cassandra and ScyllaDB that wraps a modern C/C++ driver to provide shard-aware, high-performance database access.

Provides a Cython wrapper around Mojang's modified LevelDB library for key-value storage operations in Python.

Converts geometries between WKB/EWKB, WKT/EWKT, and GeoJSON formats with no external dependencies, designed for use in PostgreSQL PL/Python functions to work with PostGIS geometry objects.

Parse, format, validate, and tokenize SQLite SQL using SQLite's own grammar and tokenizer, with full AST support and schema-aware analysis including column lineage.

Embeds a PostgreSQL server with pgvector extension into your Python application as pip-installable binaries, eliminating the need for users to install or manage PostgreSQL separately.

PyGreSQL is a Python interface to PostgreSQL databases that wraps the libpq C API, allowing direct database connections and queries from Python code.

dbt-glue is a dbt adapter that enables running dbt transformations against AWS Glue's Spark engine using the Glue Interactive Sessions API.

However, verify that your use case aligns with Spark's capabilities and Glue's cost model, and confirm production-readiness expectations given the Beta status.

PyORC reads and writes Apache ORC files using Python, wrapping the Apache ORC C++ API and providing an interface similar to Python's csv module.

Python client for Upstash Vector, a serverless vector database—upserts and queries vectors with optional metadata and data fields, supporting dense, sparse, and hybrid index types.

However, the aging maintenance status (533 days since last release) and Alpha classification warrant caution for mission-critical production systems; verify that your…

pytd provides Python interfaces to Treasure Data's REST APIs, Presto query engine, and Plazma storage, enabling efficient read/write operations on large data volumes through pandas DataFrames.

Install it if you need to query Treasure Data or write pandas DataFrames to it; use td-client-python instead if you only need basic REST API operations.

Provides transparent persistence for Python objects, allowing them to be automatically saved to and loaded from a database like ZODB without explicit serialization code.

Not recommended as a standalone persistence solution outside the ZODB ecosystem.

BTrees provides persistent, scalable B-tree data structures optimized for ZODB's optimistic concurrency model, with built-in conflict resolution for concurrent access.

A convenience wrapper around the official Redis Python client that adds automatic JSON serialization, context manager support, and helper methods for working with multiple keys at once.

No, not for new projects.

Integrates SQLAlchemy sessions with Zope's transaction manager, allowing database operations to participate in Zope's distributed transaction protocol.

Install only if you understand both SQLAlchemy and Zope transactions; it is not a general-purpose ORM integration and adds no abstraction layer.

Rust-compiled extensions that accelerate the official Neo4j Python driver by optimizing performance-critical operations, requiring no code changes to existing driver usage.

A Python client for streaming data ingestion into Databricks Delta tables via the Zerobus service, supporting both JSON and Protocol Buffer serialization with synchronous and asynchronous APIs.

asyncmy2 is a fast asyncio-based MySQL/MariaDB driver that rewrites the core protocol in Cython for performance while maintaining API compatibility with aiomysql.

Not recommended if you need Python versions below 3.10 or want to avoid build toolchain setup.

Provides a Python interface to construct, manipulate, and serialize Substrait Plans—cross-language specifications for data compute operations—for consumption by engines like DuckDB or DataFusion.

A Python client for connecting to Apache Hive Metastore and executing DDL commands via the Thrift protocol.

However, verify the license terms before use in commercial projects, and note that the latest release is over 4 years old—test compatibility with your Hive version…

A lightweight vector database that stores and queries embeddings with a single numpy dependency, designed for prototyping and small-to-medium workloads.

However, do not use for production systems: the aging maintenance status (last commit 2026-01-09, first release 2024-08-19), lack of active development, and…

Implements the Ethereum Merkle Trie data structure for efficient key-value storage and retrieval with cryptographic root hashing.

Install it if you are working with Ethereum state structures, building blockchain tooling, or need a production-grade Merkle Trie implementation in Python.

Manages a pool of reusable PostgreSQL connections via psycopg2, reducing connection overhead in multi-threaded applications.