Weaviate for vector databases
Weaviate is a vector database with hybrid search built in, and its MCP server lets an agent query and manage one directly. It ranks third of four for vector databases, which fits its specialty: it is the pick when you want to combine semantic similarity with keyword matching in a single query, rather than the broadest general-purpose store.
The seed names the edge precisely. Weaviate's hybrid search blends vector similarity with BM25 keyword matching, so retrieval that needs both meaning and exact terms is where it stands out. It earns the middle of the pack as the hybrid-retrieval specialist among more general vector stores.
How Weaviate fits
The retrieval tool is weaviate-query-hybrid, which runs a search combining vector similarity with keyword (BM25) matching over a collection, which is the capability that makes Weaviate distinct for this task. weaviate-objects-upsert creates or updates objects so an agent can index documents into a collection, and weaviate-collections-get-config inspects a collection's schema, properties, and vectorizer settings before querying. For multi-tenant setups, weaviate-tenants-list enumerates the tenants in a collection.
The honest scope: this is a focused tool set built around hybrid query, upsert, and schema inspection, not a full management console. Pinecone is the fully managed cloud vector service when you want minimal operations, Qdrant is the self-hostable open-source engine, and Chroma is the embedded library for local use. Choose Weaviate when hybrid semantic-plus-keyword retrieval is the requirement and an agent should run those queries and index objects itself.
Tools you would use
| Tool | What it does |
|---|---|
| weaviate-collections-get-config | Inspects a collection's configuration, including its schema, properties, and vectorizer settings. |
| weaviate-tenants-list | Lists the tenants within a multi-tenant collection. |
| weaviate-query-hybrid | Runs a hybrid search that combines vector similarity with keyword (BM25) matching over a collection. |
| weaviate-objects-upsert | Creates or updates objects in a collection. |
FAQ
- What makes Weaviate's server different for vector search?
- Hybrid search. weaviate-query-hybrid combines vector similarity with BM25 keyword matching in one query, so retrieval that needs both semantic meaning and exact terms is where Weaviate stands out among the picks.
- Can the agent index documents and inspect the schema?
- Yes. weaviate-objects-upsert creates or updates objects in a collection, and weaviate-collections-get-config returns the schema, properties, and vectorizer settings, so an agent can index data and check a collection's setup before querying.