OpenVikingStore is a LangGraph BaseStore backed by OpenViking. Use it for long-term memory across threads in an OpenViking HTTP deployment.
Overview
OpenViking is a context database for AI agents. Thelangchain-openviking package includes an OpenVikingStore class for storing LangGraph values as durable OpenViking content records and indexing them for query-based search calls.
This page documents the published langchain-openviking package on PyPI (v0.1.0).
Integration details
Setup
To useOpenVikingStore, you need an OpenViking HTTP deployment and an API key.
Installation
Install the integration with thelanggraph extra:
Credentials
Set your OpenViking URL and API key:Instantiation
InstantiateOpenVikingStore with your OpenViking HTTP connection settings:
root_uri is viking://user/memories/langgraph_store. Newer OpenViking servers reject that uid-less form, so pass the viking://~ home alias (as above) or an explicit viking://user/<uid>/... URI.
Usage
Useput to write a JSON-serializable mapping under a namespace and key:
get to retrieve the stored item:
search with a namespace prefix. Pass query for semantic search:
delete to remove a key:
API reference
For detailed documentation, see the OpenViking LangChain and LangGraph integration guide and theOpenVikingStore source for v0.1.0.
See also the LangGraph store integrations list.
Connect these docs to your agent of choice via MCP for real-time answers.

