RuVector

RuVector is a Rust native vector search and agent memory substrate for fast local decisions across sessions.

RuVector is a Rust native substrate for fast local decisions and agent memory across sessions. It combines local semantic embeddings, persistent vector retrieval, graph relationships, explicit feedback learning, memory lifecycle controls, and optional shared memory. It supports three complementary roles including sense and respond, learn and remember, and reason and orchestrate. The software is designed for developers building local AI workflows, agent memory systems, and typed local decisions. Users can install it via npm, PyPI, or directly from crates.io. It integrates with Node.js, Python, Rust, browser WASM, and Model Context Protocol clients.

RuVector at a glance

Best for: Software engineer, Researcher
Runs on: Web, Node.js, Python, Rust, CLI, API, self-hosted, open source
Pricing: Free
Released: October 2026
Category: AI search engine
Website: github.com/ruvnet/ruvector

Key features

Vector search and persistent vector retrieval
Graph relationships and memory
Local semantic embeddings via ONNX
Explicit feedback learning and SONA
Typed local decisions and classification
Browser WebAssembly support
Model Context Protocol server integration

Use it for

Search vectors and retain agent context
Classify text or return typed local decisions
Route support tickets locally
Build knowledge graph retrieval systems

Alternatives to RuVector

HelixDB — Combines graph, vector, and full-text search in one engine.
LEANN — Efficiently index millions of documents locally with 97% storage savings.
ColdIVF — It reads under 1 percent of the index during cold queries.
TurboVec — Fits a 10 million document corpus in 4 GB of RAM and searches faster than FAISS.
zvec — It is a pure local, serverless vector database providing fast hybrid search.
SwarmVault — Compiles raw sources into a local-first knowledge graph without requiring cloud access.
SeekStorm — It delivers sub-millisecond lexical and vector search performance in Rust.
Qdrant — Fast filtered vector search, self-hosted or in a managed cloud
Exa — Provides high-accuracy web search and deep research endpoints for AI agents.
Open Glean — Provides agents with structured state and relationship-based context rather than just similarity.
Vespa — Combines vector database capabilities with distributed machine-learned model inference.
zvec-grep — Unifies ripgrep, BM25, and vector search behind one local-first interface.

All RuVector alternatives