oceanbase/seekdb

What is it?

Key features

• 01Vector, full-text and scalar filters combined in one SQL query
• 02MySQL protocol compatibility with full ACID transactions
• 03Embedded mode for Python and Node.js, plus single-node server deployment
• 04FORK DATABASE copy-on-write snapshots and MERGE TABLE for agent sandboxes
• 05Asynchronous index pipeline with incremental HNSW for near-real-time vector search

Use cases

• Persistent memory and state storage for AI agents
• Retrieval-augmented generation (RAG) backends that need vector and keyword search together
• Sandboxed experimentation on agent data with rollback or merge
• Adding semantic search to an existing MySQL-based application

Quick start

1. 1

   Install the Python packages (requires Python 3.11 or newer)

   python3 -m pip install -U pylibseekdb pymysql
2. 2

   Save the README example as example.py

3. 3

   Run the example; it should print [{'value': 1}]

   python3 example.py

Things to know

• !The embedded Python quick start requires Python 3.11 or newer.
• !Supported platforms are limited to those listed on the pylibseekdb package page.

Topics

FAQ

What is seekdb?

An AI-native search database that stores vector, full-text, structured and semi-structured data in one MySQL-compatible engine, for agent memory and retrieval. seekdb is a database built for AI agent workloads. It combines vector search, full-text search and scalar filters in a single engine and exposes them through SQL, so retrieval does not need separate systems or client-side result merging. It is MySQL-compatible and offers full ACID transactions, and it can run as an embedded library from Python or Node.js, as a single-node server, or as part of an OceanBase distributed cluster.

Is seekdb free to use? What license does it have?

seekdb is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.

Is seekdb still maintained?

Actively maintained. The last commit to seekdb was on Oct 21, 2026.

What can I use seekdb for?

Persistent memory and state storage for AI agents Retrieval-augmented generation (RAG) backends that need vector and keyword search together Sandboxed experimentation on agent data with rollback or merge Adding semantic search to an existing MySQL-based application

How do I install seekdb?

Install the Python packages (requires Python 3.11 or newer): python3 -m pip install -U pylibseekdb pymysql Save the README example as example.py Run the example; it should print [{'value': 1}]: python3 example.py

Where can I get seekdb?

The source code and installation instructions are on GitHub: https://github.com/oceanbase/seekdb