Best MCP servers for vector databases

Vector databases store embeddings and answer similarity queries, the backbone of semantic search and retrieval-augmented generation, and an AI agent that can talk to one directly can index documents, run nearest-neighbor searches, and manage collections as part of its own reasoning loop. Rather than a separate retrieval service the agent calls blindly, an MCP connection lets the agent query the store, inspect what is there, and refine its search. The right server depends on your stack: a managed cloud vector service, a self-hostable open-source engine, or an embedded library for local use. The servers below are real MCP servers for popular vector databases, each with a verified, current install config, so an agent can run vector search directly.

Top pick

Pinecone

Pinecone

Official

Pinecone's official developer MCP server: search indexes, manage records, rerank results, and look up Pinecone docs from your agent.

vector-search

Pinecone's server lets an agent query and manage a fully managed cloud vector database, the default for teams who want production-grade similarity search without running infrastructure.

Pick 2

Qdrant

Qdrant

Official

Qdrant's official MCP server: a semantic memory layer that stores and retrieves information from a Qdrant vector database.

vector-search

Qdrant's server connects an agent to a popular open-source vector engine with rich filtering, strong for self-hosted semantic search with metadata constraints.

Pick 3

Weaviate

Weaviate

Official

Weaviate's built-in MCP server: hybrid vector and keyword search, schema inspection, and object upserts against a Weaviate vector database.

vector-search

Weaviate's server exposes a vector database with built-in hybrid search and modules, ideal when an agent needs to combine semantic and keyword retrieval.

Pick 4

Chroma

Chroma

Official

Chroma's official MCP server: manage collections and run semantic, metadata, and full-text search over a Chroma vector database.

vector-search

Chroma's server suits lightweight, developer-friendly vector storage, a good fit for prototyping retrieval and embedding-backed memory locally before scaling up.