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

Vectorrag.net

VectorRAG.Net is a C# vector database library that runs inside .NET 8 applications and supports semantic search, RAG retrieval, hybrid vector-plus-BM25 search, document chunking, metadata filtering and snapshot save/load.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Vectorrag.net — 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 it when building a .NET 8 application that needs local, in-process vector retrieval, especially where a separate vector database service is not wanted.

Who it is for

  • .NET developers adding retrieval to a RAG application or local assistant.
  • Teams building searchable internal documentation or support knowledge bases.
  • Developers who want vector search embedded in an application process rather than in a separate vector service.

Documented uses

  • Store text documents and their embeddings for application retrieval.
  • Find relevant documents by semantic similarity or combined vector and BM25 search.
  • Filter search results using document metadata.
  • Save or load a database snapshot.
Before you choose

Requirements & limits

The repository documents a .NET 8 NuGet package and lists Windows, Linux and macOS. Its README says production use requires a commercial license and quotes $300 USD as a one-time payment for a perpetual license.

  • The library runs in the application process and does not require a separate HTTP or Docker-based vector service.
  • The application must supply an embedding model; the README's document-ingestion example passes one to the library.
  • The README's quick-start example requires a license key.
  • Search supports vector retrieval and an optional vector-plus-BM25 hybrid mode, as well as metadata filters.
  • Published benchmark figures are tied to a local test setup; the README says performance varies with hardware, dataset size and embedding dimensions.

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

Set up VectorRAG.Net for semantic document search

Add text documents to an in-process .NET 8 vector database and retrieve the top five results for a semantic query.

Before you start

  • A .NET 8 application.
  • The VectorRAG.Net package, added with dotnet add package VectorRAG.Net.
  • A license key for database initialization.
  • An embedding model that supports document embedding and query embedding.
  • A text file such as faq.txt.

Steps

  1. Add the package with dotnet add package VectorRAG.Net.

  2. Initialize VectorRAGDatabase with the embedding dimension, an EmbeddingLshConfig, and the license key.

  3. Call UpsertTextDocumentAsync with an external ID, the text file contents, optional metadata, and the embedding model.

  4. Generate a query embedding with the embedding model, then call db.Search with SearchOptions { TopK = 5 }.

Expected outcome

The database contains the added text document and returns up to five search results for the embedded query.

Guide limitations

  • The example requires a license key; production use requires a commercial license.
  • The example assumes an embedding model is available; its setup is not described.
  • Actual performance depends on embedding dimensions, dataset size, and hardware.
Documentation 1 ↗
Source-based practical guide

Save and reload a VectorRAG.Net database snapshot

Persist an initialized VectorRAG.Net database to a snapshot file and load it again.

Before you start

  • A .NET 8 application with VectorRAG.Net installed.
  • An initialized VectorRAGDatabase instance.

Steps

  1. Call await db.SaveAsync("database.vdb") to save the database snapshot.

  2. Call await db.LoadAsync("database.vdb") to load the snapshot.

Expected outcome

A snapshot is saved to, and can be loaded from, `database.vdb`.

Guide limitations

  • The documented calls do not describe snapshot compatibility, error handling, or whether loading merges with or replaces existing data.
  • The database initialization and license requirements still apply.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
likeslines-maker/VectorRAG.Net · Official homepage title · retained in its source language
How we review sourcesSuggest a correction