stoolap/stoolap

Health v2 · maintenance only

Adoption not part of the score

1210 stars · 45 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Stoolap is a modern embedded SQL database written in pure Rust, offering ACID transactions with MVCC, a cost-based query optimizer, parallel query execution, and native vector search via HNSW indexes. It runs in-process with no external server and ships first-party drivers for Python, Go, Node.js, Java, C#, and other languages.

Use cases

• embed a SQL database in a Rust application without an external server
• run transactional workloads with MVCC and snapshot isolation
• query historical data with AS OF time-travel queries
• add vector similarity search to an embedded database
• use SQL from Python, Go, Node.js, Java, or C# via native drivers
• run analytical queries with parallel execution and columnar storage

When to choose

• you want an embedded, in-process SQL database with modern features like MVCC and time-travel queries
• you need vector search alongside relational SQL in a single binary
• you prefer a pure Rust, memory-safe alternative to SQLite or DuckDB for embedded use
• you need multi-language drivers for an embedded engine

When to avoid

• you need a battle-tested database with a long production track record
• you require a client-server database with network access from many hosts
• your stack depends on mature ecosystem tooling and extensions built for SQLite or PostgreSQL

Facets

library · maturity active

database search-engine vector-database databases developer-tools backend rust cross-platform python go jvm dotnet wasm embedded-database sql mvcc acid time-travel-queries hnsw vector-search columnar-storage query-optimizer in-memory-database nodejs

Member repositories

For agents

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