ruvnet/RuVector

Health v2 · maintenance only

Adoption not part of the score

4456 stars · 589 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

RuVector is a Rust-native vector database and memory substrate for AI agents, combining persistent semantic embeddings, graph relationships, and feedback-based learning. It ships as a Rust crate and an npm package with a CLI for storing and recalling agent memories locally without a database server.

Use cases

• give ai agents persistent memory across sessions
• store and recall semantic memories locally
• build a vector database in rust
• add graph relationships to vector retrieval
• run local embeddings without an api key
• record feedback so agent memory improves over time

When to choose

• you need persistent, self-learning memory for AI agents with local-first storage
• you want a high-performance Rust vector database with graph and GNN features
• you need semantic recall without running a database server or external API

When to avoid

• you need a battle-tested, widely adopted vector database like pgvector or Qdrant for production workloads
• you require multi-model or multi-dimension embedding stores without reembedding
• you need managed hosting and a mature ecosystem around your vector store

Facets

library · maturity active

vector-database rag machine-learning database search-engine cli databases artificial-intelligence developer-tools rust cli cross-platform wasm agent-memory embeddings graph-neural-networks onnx persistent-memory self-learning low-latency ai-agents retrieval-augmented-generation nodejs

Member repositories

For agents

markdown · JSON · MCP: product_card(name="ruvnet/RuVector")