SQLite is evolving into a modern data layer. The star is SQLite Graph Ext, an alpha extension that adds graph capabilities and Cypher queries to SQLite. Built by Agentflare AI, it's MIT licensed and aims to run end-to-end patterns like CREATE, MATCH, and RETURN on a single SQLite engine [1].
Alpha caveats aside, the project reports strong throughput—340K nodes/sec inserts, 390K edges/sec, and 180K node scans—with full CREATE and MATCH support and a virtual table bridge to mix SQL and Cypher [1]. Limitations linger: only forward relationships, no relationship property filtering, no aggregations or bidirectional paths yet [1].
Meanwhile, the broader trend toward offline-first data layers is on display in browser-based personal databases. Superego is an open-source, local database with an AI assistant, designed to store essentially anything about your life. It runs entirely in your browser, with no server or login, emphasizing data ownership [2].
It also supports building tiny personal apps that use the data, with optional offline inference via Ollama/LM Studio when you have the horsepower, or a remote API if you don’t [2].
Another offline-friendly move? hosting SQLite databases on GitHub Pages—a simple way to serve local DBs for offline use in a browser [3].
Taken together, SQLite is stepping up as an edge-friendly data layer that blends graph patterns, on-device AI, and offline hosting.
References
Show HN: SQLite Graph Ext – Graph database with Cypher queries (alpha)
SQLite extension adds graph capabilities via Cypher; creates nodes/edges, queries with MATCH; discusses pipeline, limits, and comparisons to Kuzu/Cozodb.
View sourceLocal, browser-based personal database with AI assistant; data ownership, connectors, and plans for multimodal data and apps offline no servers.
View sourceHosting SQLite Databases on GitHub Pages
Discusses hosting an SQLite database via GitHub Pages, outlining limitations, approaches, and potential use cases, security considerations, and offline capabilities.
View source