Weaviate
Open-source vector database with hybrid search built in.
- made by
- Weaviate
- source
- Official Weaviate site
Magic Ship is one shop in Vancouver, BC, working remotely with clients worldwide. We are not a partner, reseller, or certified vendor of Weaviate - we just build with this.
What Weaviate is
Weaviate is an open-source vector database with a schema, REST and GraphQL APIs, and hybrid search that fuses BM25 and vector results natively. Modules can generate embeddings and rerank at query time, multi-tenancy is a first-class feature, and it runs self-hosted or as a managed cloud.
How we use it
We use it when hybrid search and reranking should live in the store rather than in application code, and when strict per-tenant isolation is a requirement rather than a filter. Multi-tenant collections let inactive tenants be offloaded, which keeps memory sane across a long tail of small customers. Schema and index settings stay in code so an environment can be rebuilt from scratch.
Where it is the wrong choice
It is another distributed system to operate, with its own memory profile and upgrade path, and its module system pulls embedding choices into database configuration where they are less visible to the people reading the application. When the vectors belong beside relational data, pgvector is simpler.
Service lines it turns up in
Related tools
More in Retrieval and vector
Other tools in the same service lines
Building something on Weaviate?
Send the problem rather than a job spec. You get an answer on scope, on fit, and on whetherWeaviate is even the right call for it.
Start a projectWeaviate and Weaviate are trademarks of their respective owners, used here to say what we work with.