All Tools
L
DataFreemiumOpen Source
LIBSQL
Open-source SQLite fork with vector search and replication
MIT
ABOUT
SQLite is the world's most deployed database but lacks features critical for AI applications — no vector search for embeddings, no native replication, and no client-server protocol. Building AI features on SQLite requires bolting on vector indexing, replication layers, and custom networking. libSQL extends SQLite with vector search (cosine, Euclidean, dot product), WASM-based user-defined functions, snapshot-based replication, and a standalone server mode while maintaining full SQLite compatibility.
INSTALL
pip install libsql-clientINTEGRATION GUIDE
1. Store and query AI embeddings alongside application data in a single embedded database
2. Run a local vector database with SQLite compatibility for edge and mobile AI apps
3. Deploy a replicated database cluster that syncs to on-device SQLite for offline-first AI
4. Use libSQL as a drop-in replacement for SQLite with added vector search capabilities
5. Build serverless AI backends with Turso's managed libSQL platform
TAGS
databasesqlitevector-searchembeddedserverlessreplication