🗄️ Free & Open-Source NoSQL Software
Key-Value & In-Memory Stores
- Redis (Community/Forks): While Redis itself shifted to non-open source licensing, the industry standard for a fully free, BSD-licensed alternative is now Valkey or KeyDB. They are ultra-fast, in-memory key-value caches and data structures.
👉 Official Valkey Project Website / Official KeyDB Website [1, 2, 3, 4, 5] - RocksDB: An embeddable, persistent key-value store for fast storage written by Meta. It is widely used as the underlying storage engine for many other larger database systems.
👉 Official RocksDB Website [1, 2, 3]
Document Stores
- Apache CouchDB: A fully free, open-source NoSQL document database that uses JSON to store data, JavaScript for MapReduce indexes, and HTTP for an API. It features seamless multi-master syncing.
👉 Official Apache CouchDB Website [1, 2, 3, 4, 5] - FerretDB: Since MongoDB went proprietary (SSPL), FerretDB was created as a 100% open-source (Apache 2.0) replacement. It serves as a translation layer that allows you to run MongoDB workloads directly on top of PostgreSQL.
👉 Official FerretDB Website [1, 2, 3]
Wide-Column Stores
- Apache Cassandra: A highly scalable, distributed NoSQL database designed to handle massive amounts of structured data across many commodity servers with no single point of failure.
👉 Official Apache Cassandra Website [1, 2, 3, 4] - ScyllaDB (Open Source Edition): A drop-in Cassandra alternative rewritten from Java into C++. It provides massive throughput and single-digit millisecond latencies while running on significantly less hardware.
👉 Official ScyllaDB Website [1, 2, 3, 4, 5]
Research NoSQL Database Projects on GitHub
SurrealDB: An innovative multi-model cloud-native database. It acts as a document, graph, and relational database all in one engine, optimized entirely for real-time web applications and AI agent memory environments.
👉 SurrealDB GitHub Repository
Meilisearch: A next-generation, AI-powered hybrid search engine. Written completely in Rust, it acts as a highly optimized document-store index tailored for lightning-fast search-as-you-type web integrations.
👉 Meilisearch GitHub Repository
BadgerDB: An influential research-backed key-value store written entirely in Go by Dgraph Labs. It separates keys from values (based on the WiscKey academic paper) to dramatically minimize write amplification compared to traditional LSM-tree engines.
👉 BadgerDB GitHub Repository
MapDB: An embedded Java database engine that provides concurrent Maps, Sets, and Queues backed by off-heap memory or disk storage. It is heavily used in Java system research.
👉 MapDB GitHub Repository
ZenFS (RocksDB Plugin): A Western Digital research initiative that integrates RocksDB directly with modern Zoned Namespaces (ZNS) SSDs, bypassing the traditional Linux file system layer to dramatically optimize flash memory endurance.
👉 ZenFS GitHub Repository [1, 2, 3, 4, 5, 6, 7, 8]