After 25 pre-production releases and more than 1,750 commits, DefraDB has reached its first stable release candidate, leading up to our v1.0 release.
This release represents a significant body of work from a team that spans the globe, shaped by deep technical experimentation, early partner feedback, and a clear view of what the future at the edge looks like.
v1.0 RC1 is a signal: the API is locked and DefraDB is ready for production workloads. If you've been watching the project from a distance, this is the version to take seriously.
Before we get into what shipped, it's worth stepping back. The question most people ask isn't "what version is this?" It's "why does this exist?"
The short answer
Most databases assume your data lives in one place. DefraDB doesn't.
DefraDB is a peer-to-peer document database designed for environments where data is created, stored, and queried across multiple devices and locations. No central server keeps things in sync. The nodes handle that themselves.
This matters now because AI is moving to the edge. Models running on devices need fast local access to structured data for inference, context, and personalization. That data needs to sync across a fleet of devices, stay consistent during disconnections, and do all of this without routing through the cloud. DefraDB was built for this.
It runs on edge devices, on-prem servers, laptops, and in the cloud. It syncs across all of them without forcing you to pick between local performance and distributed availability.
How it works
DefraDB combines a few ideas that make this possible.
Documents, not rows. DefraDB is a NoSQL document store. You define schemas, store JSON-like documents, and query them with DQL, a query language compatible with GraphQL. If you've used MongoDB or Firestore, the interaction model will feel familiar.
Merkle CRDTs for conflict-free sync. Every document update is stored as a node in a Merkle DAG, using Conflict-Free Replicated Data Types. Two devices can independently modify the same document offline, and when they reconnect, the changes merge deterministically. No conflict resolution logic required. The merge result is identical regardless of the order updates arrive.
Content-addressable data. Every piece of data in DefraDB is identified by its content, not its location. Any node on the network can serve a request and the recipient can cryptographically verify the response. No trust assumptions about where data came from.
Peer-to-peer networking. Nodes discover and communicate directly with each other. There's no central coordination server. You can set up passive sync (pubsub), active replication (push changes to specific peers), or both.
Automatic schema transformation. In edge environments, not every device runs the same software version. Schemas drift. DefraDB includes a built-in data transformation engine (LensVM) that uses bi-directional lenses to convert data between schema versions on the fly. Devices running v3 of your app can sync with devices still on v2 without requiring a coordinated update or a centralized schema registry. The transformations are defined once, composed into chains, and executed locally at the database layer.
Built-in access control. DefraDB supports relationship-based access control at the document level, using cryptographic identity (DIDs), with field-level control coming soon. Policies travel with the data, not bolted on at the application layer.
What v1.0 RC1 means
This release is the result of a focused push toward API stability. Here's what changed.
Stable public API across every surface. The Go SDK, CLI, HTTP/OpenAPI endpoints, GraphQL schema, C bindings, and JavaScript bindings all went through a systematic naming audit. Verb-first conventions, consistent terminology, typed option builders. The full rename changelog is in the release notes. It's extensive, and it's intentional. When we ship v1.0 final, these interfaces won't break under you.
New capabilities. This RC introduces branchable collection sync, searchable encryption, LevelDB as a storage backend for WASM deployments, secp256k1 public key support, a setup wizard for first-run configuration, collection truncation, and recursive explain for nested query joins. Node Access Control now gates every operation in the system. Alongside these features, there's a substantial round of bug fixes and comprehensive extensions to the integration test suite.
Cross-platform reach. DefraDB compiles to a single binary. With the C bindings and JS bindings that shipped in the v0.18-v0.20 cycle (and are now stabilized), you can embed DefraDB into mobile apps, browser-based applications, IoT firmware, or virtually any runtime supporting WASM or C-interoperability.
Who this is for
Edge AI is the most immediate use case. Models running on devices need fast local access to structured data for inference, context, and personalization. That data needs to sync across devices, update without cloud round-trips, and stay consistent even when connectivity is intermittent. DefraDB was designed for exactly this: it gives on-device AI a local, queryable, self-syncing data layer that works whether the device is online or not.
Beyond AI, DefraDB fits anywhere data needs to work at the edge: robotics, autonomous systems, field operations, collaborative tools. If your application can't depend on a round-trip to the cloud for every read or write, DefraDB is the data layer for that.
Try it
Install from source or grab a binary from the release:
The full changelog and migration notes from v0.20.0 are in the release on GitHub.
This is a release candidate. If something breaks or an API feels wrong, now is the time to tell us. File issues, open discussions, or find us on Discord.
v1.0 final is next.
