Introduction

FlintDB is a modern embedded analytical database designed for applications that need local-first data storage with powerful analytics — without the complexity of SQL.

Key Concepts

Query is Data

In FlintDB, queries are plain JSON objects. There is no SQL parser, no query language to learn. Your query is just data that describes what you want.

{
  "collection": "sales",
  "filter": { "region": "US" },
  "groupBy": "category",
  "aggregate": { "total": { "sum": "amount" } },
  "sort": { "total": "desc" },
  "limit": 10
}

Flexible Input, Strict Output

Store any JSON document without defining a schema upfront. FlintDB applies types at query time, giving you structured, typed results from schema-free storage.

Two Access Modes

Mode	Latency	Use Case
Native (napi-rs)	~2 us/get	In-process, sync, maximum performance
IPC (Unix socket)	~50 us/get	Multi-process, async, daemon-based

Both modes share the same Rust engine underneath.

Architecture

FlintDB is built in Rust with:

• redb — mmap-backed, ACID, crash-safe storage via Copy-on-Write
• Lazy loading — collections deserialized on first access
• Columnar cache — dense f64 arrays built lazily per field for fast analytics
• Hash & BTree indexes — auto-maintained on mutations