LanceDB vs Milvus and Zilliz Cloud API + MCP

LanceDB

Embedded, no server, and tables can live directly in S3, GCS or Azure storage. No self-serve hosted API. Enterprise is sold through a contact form.

Milvus and Zilliz Cloud API + MCP

Apache-2.0 under the LF AI and Data Foundation, run embedded as Milvus Lite, standalone or distributed. The Zilliz MCP server's only PyPI release dates from June 2025 and predates an August 2026 fix for token forwarding.

LanceDB

1. Use it through the Python or TypeScript library. There's no public endpoint an agent can call without an Enterprise deployment
2. Build an FTS index before calling query_type="hybrid", or hybrid search has no keyword side
3. Call optimize() after many small writes, since each write adds a new file version
4. Use select() to drop the vector column from results unless the task needs it
5. On object storage, set a read consistency interval if several processes write the same table

Milvus and Zilliz Cloud API + MCP

1. Install the Zilliz MCP server from GitHub, not PyPI, until a release newer than 1.0.0 ships
2. Use the cluster endpoint for data (/v2/vectordb/entities/search) and api.cloud.zilliz.com only for cluster management
3. Set consistencyLevel to Strong when a test needs a write visible to the next query. The default is Bounded
4. Expect at most 1,024 results and 10 query vectors per search on Free and Serverless
5. Return only the fields you need. Returning the vector field multiplies read cost