# cozodb/cozo

A transactional, relational-graph-vector database that uses Datalog for query. The hippocampus for AI!

Repository: https://github.com/cozodb/cozo
Canonical: https://ross.abutalabs.com/products/cozo
Homepage: https://cozodb.org
Language: Rust
License: MPL-2.0
License Family: copyleft
Topics: database, graph, client-server, cross-platform, datalog, embedded-database, graph-algorithms, graph-database, graphdb, relational-database, single-executable, cozo, cozoscript
Last push: 2024-12-04T12:49:06+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 1419, "days_push": 637, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4097, forks 164 (observed 2026-08-28T04:08:35.431117+00:00)

## What it is
CozoDB is a transactional, relational-graph-vector database written in Rust that uses Datalog as its query language. It runs embedded or as a client-server service, supports HNSW vector search, full-text search, and graph algorithms, with bindings for many languages.

## Use cases
- store and query graph data with Datalog
- run vector similarity search alongside relational joins
- embed a lightweight database in an app with a single executable
- run graph algorithms like PageRank on stored relations
- build RAG applications needing vector search plus graph traversal
- find near-duplicate documents with MinHash-LSH
- query recursive relationships in connected data

## When to choose
- you want Datalog's declarative recursive queries over graph data
- you need vector search integrated with relational and graph queries in one engine
- you want an embeddable, cross-platform database with bindings for many languages

## When to avoid
- you need a mature, widely-supported SQL database with a large ecosystem
- you require heavy multi-node horizontal scaling or enterprise-grade replication
- your workload is simple key-value storage without graph or vector needs

## Facets
- artifact type: library
- maturity: active
- function: database, vector-database, search-engine
- domain: databases, graph-processing, artificial-intelligence
- platform: cross-platform, rust, python, wasm, jvm, go, embedded
- tags: datalog, graph-database, embedded-database, hnsw, vector-search, full-text-search, minhash-lsh, single-executable, search, nodejs, swift

## Member repositories
- cozodb/cozo (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:35.431117+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:23:18.408079+00:00, confidence not recorded.
  - readme: https://github.com/cozodb/cozo (fetched 2026-08-28T04:08:35.431117+00:00, sha e436a4daae09)
  - homepage: https://cozodb.org (fetched 2026-08-29T09:14:54.339282+00:00, sha ad19ac36e282)
  - registry_crates: https://crates.io/api/v1/crates/cozo (fetched 2026-08-29T09:14:54.348219+00:00, sha 15cc75a61227)
- Data as of 2026-08-30T08:39:29.467469+00:00.