OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
RAG & Knowledge Systems

Lynsedb

LynseDB is a Python-facing vector database with a Rust storage and search engine. It supports embedded use, an HTTP service and a lightweight sharded cluster, with retrieval features for RAG, semantic search, agent memory and other similarity-search tasks.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Lynsedb — OpenViST project profile in RAG & Knowledge Systems.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use LynseDB when a Python application needs vector or similarity retrieval and you want a documented path from a local embedded database to shared HTTP access or a lightweight sharded cluster.

Who it is for

  • Python developers adding retrieval to AI applications
  • Teams building RAG or document question-answering systems
  • Developers implementing semantic search, agent memory or recommendation features
  • Applications that need to move from single-process embedded storage to shared service or cluster access

Documented uses

  • Store vectors alongside stable IDs and metadata, then retrieve relevant records by similarity.
  • Filter or combine retrieval using metadata, BM25, dense or sparse vectors, and hybrid search.
  • Return stored fields as grounded context for an LLM.
  • Search non-embedding data such as coordinates, binary fingerprints, numeric profiles or probability distributions.
Before you choose

Requirements & limits

The repository is public and identifies the project as Apache-2.0 licensed. Python 3.9 or newer is required. The README lists Linux, macOS and Windows native support.

  • Embedded mode is for one Python process owning the data directory; use the HTTP service when independent processes need shared access.
  • The same Python client is documented for embedded, HTTP service and cluster modes.
  • Document-based insertion and search can use the optional local embeddings adapter; existing vectors can instead be passed directly.
  • The README distinguishes logical commits from checkpoints and recommends checkpoints for specified durability-sensitive operations.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Build a small document retrieval collection

Store knowledge snippets with metadata and retrieve a matching result using a question and metadata filter.

Before you start

  • Python 3.9 or newer
  • The LynseDB Python package
  • Install the optional embeddings adapter with pip install "lynsedb[embeddings]" when using document-based insertion and search with local embeddings

Steps

  1. Install LynseDB and, for local document embedding, its optional embeddings adapter.

  2. Create a VectorDBClient with a local data directory, then create a database and collection.

  3. Add documents with stable IDs and metadata fields inside the collection context.

  4. Search with a question, a result count and a metadata filter; request returned fields and inspect the results.

Expected outcome

A local collection returns matching stored fields that can be used as context for an LLM.

Guide limitations

  • The example uses local embedded storage; multiple independent processes should not share the same embedded data path.
  • For production, choose and version the embedding model, or pass vectors directly if embeddings are already generated.
  • The README recommends checkpointing before backups, snapshots, controlled shutdowns or critical durability acknowledgements.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Vector search

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
LynsedbStore vectors alongside stable IDs and metadata, then retrieve relevant records by similarity.Cloud, Linux, macOS, Self-hostedThe repository is public and identifies the project as Apache-2.0 licensed. Python 3.9 or newer is required. The README lists Linux, macOS and Windows native support.
MettisIndex text documents through an HTTP endpoint.Linux, Self-hosted, Server
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
BirchKwok/lynsedb · Official homepage title · retained in its source language
How we review sourcesSuggest a correction