Weaviate
An open source vector database in Go that combines vector and keyword search, with built-in embedding modules. See how developers rate it month by month on DTDb.
- Category
- database engine
- Developer
- Weaviate
- Created by
- Bob van Luijt, Etienne Dilocker
- First release
- 2019
- License
- BSD-3-Clause
- Open source
- Yes
- Pricing
- Free with paid plans
- Website
- https://weaviate.io
- Written in
- Go
- Data model
- Vector, Search
- Storage layout
- Vector index
- Workload
- Search
- Analytics architecture
- Not an analytics platform
- Query language
- GraphQL, REST, gRPC
- ACID transactions
- No
- Schema
- Flexible
- Deployment
- Server, Distributed cluster, Managed cloud
- Scaling
- Sharding
- Create, read, update, delete
- Create: fast (Batch imports with asynchronous indexing); Read: fast (HNSW with hybrid keyword search); Update: moderate (A new vector re-indexes the object); Delete: moderate (Tombstones in the index, cleaned up later)