filedb vs tidesdb
Embedded key-value storage projects compared
filedb and tidesdb are storage libraries for applications that need key-value data without adopting a full database server. filedb uses append-only logs and an in-memory index in Zig, while tidesdb is a transactional C library built around an LSM-based design with column-family support.

filedb: Store Key-Value Data in Append-Only Files
FileDB is a Zig key-value store inspired by Bitcask, using append-only disk files and an in-memory index for lookups. It also includes a Redis-compatible client, but its inactive status and single-contributor history make production adoption a decision that needs careful review.

tidesdb: Transactional LSM-Tree Storage Engine in C
TidesDB is a C library for transactional key-value storage built on an LSM-tree, with a hybrid B-tree design. It is intended for embedding in applications or using as a standalone key-value or column store, and supports Linux, macOS, Windows, and other platforms.
| filedb | tidesdb | |
|---|---|---|
| Language | Zig | C |
| License | — | MPL-2.0 |
| Stars | 198 | 602 |
| Forks | 7 | 59 |
| Last analyzed | Oct 9, 2026 | Oct 22, 2026 |
Key differences
- filedb is implemented in Zig and tidesdb is implemented in C.
- filedb appends writes to data files and keeps record-location metadata in memory; tidesdb uses an LSM-tree-based design with a hybrid LSM and B-tree approach.
- filedb offers basic insert, retrieve, delete, and key-listing methods; tidesdb adds ACID transactions, MVCC, multiple isolation levels, and column families.
- filedb describes periodic syncing and an option to sync each request; tidesdb offers configurable durability from none to full sync.
- filedb is described as an early-stage project created and last pushed in June 2025; tidesdb describes cross-platform support and a broader set of storage features.
- tidesdb uses the MPL-2.0 license for the project, with additional licenses applying to dependencies; filedb is suited to Zig developers evaluating a compact storage layer.
Choose filedb if you…
- want a Zig-based key-value library with append-oriented writes and an in-memory index.
- are prototyping basic key-value access or learning about Bitcask-style storage and compaction.
- can evaluate an early-stage project and its syncing tradeoffs before relying on it for important data.
Choose tidesdb if you…
- need transactional key-value or column-family storage embedded in a C application.
- require ordered scans, range deletes, or multiple transaction isolation levels.
- want configurable compression, TTL, or durability policies in a storage engine.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.