Weaviate
Weaviate is an open source vector database by Weaviate that stores objects and supports vector, keyword, and hybrid retrieval for AI applications.
Written by
What Weaviate does
Weaviate is a vector database for storing objects and retrieving them with vector, keyword, or hybrid search. Builders can self-host Weaviate or use Weaviate Cloud, then connect model providers and application code around a database designed for AI retrieval.
Weaviate includes a Model Context Protocol server at an instance's REST API endpoint. The server can expose schema inspection, vector or hybrid search, and object updates, subject to instance authentication and authorization settings.
Who Weaviate is for
- AI applications that need vector and keyword retrieval in one database.
- Teams that want a free managed cluster for evaluation before choosing a paid cloud tier.
- Agent workflows that need an MCP connection to search or inspect a Weaviate instance.
Who should pick something else
- Teams that want to enable an unaudited write-capable agent by default, because the MCP server can modify objects.
- Large production workloads on the free Cloud tier, because it has fixed object, memory, disk, collection, and tenant limits.
- Teams that do not want to configure a self-hosted database or pay for managed capacity.
Weaviate: one strength, one limit
Strength. Weaviate Cloud's free tier includes one managed cluster per user and supports 100,000 objects.
Limit. On self-hosted deployments, Weaviate's MCP server is disabled by default for security.
What Weaviate costs
Weaviate Cloud has a permanent free cluster with fixed capacity. Paid cloud plans are billed by cluster and usage dimensions, while self-hosting shifts infrastructure operations to the team.
Current prices: weaviate.io . Prices change often, so check the maker's own page before you buy.
Weaviate with AI coding agents
Weaviate has a built-in MCP server on a database instance and a separate hosted Weaviate Docs MCP server. The database MCP server can inspect schemas, run vector or hybrid searches, and modify objects, while the docs MCP server serves development documentation.
Questions people ask
Does Weaviate offer MCP support for an AI coding environment?
Yes. Weaviate documents both an instance MCP server and a hosted documentation MCP server, using Streamable HTTP and supporting clients including Claude Code and Cursor.
Alternatives to Weaviate
- Pinecone: Pinecone is a managed vector database by Pinecone that stores and searches dense, sparse and full-text indexes for retrieval, recommendation and AI applications.
- Qdrant: Qdrant is an Apache-licensed vector search engine by Qdrant that stores dense and sparse vectors, filters by payload, supports hybrid retrieval and runs self-hosted or as managed Qdrant Cloud.
Compare Vector databases tools
All three have a free plan.
Qdrant uses the Apache-2.0 license; Weaviate uses the BSD-3-Clause license; Pinecone is proprietary.
More in Vector databases
Pinecone is a managed vector database by Pinecone that stores and searches dense, sparse and full-text indexes for retrieval, recommendation and AI applications.
Qdrant is an Apache-licensed vector search engine by Qdrant that stores dense and sparse vectors, filters by payload, supports hybrid retrieval and runs self-hosted or as managed Qdrant Cloud.