pymilvus

Why we included this project

pymilvus is the official Python client for Milvus, the widely used open-source vector database, and the piece most teams reach for when they build real vector workloads. It connects Python code to a Milvus cluster, and it also runs against Milvus Lite, the embedded version you can spin up locally. The SDK covers the full production lifecycle: creating collections, defining schemas and indexes, inserting and upserting data, and running vector and hybrid searches with tunable parameters. Optional install extras add a model package that turns text or images into embeddings and a bulk-writer utility for high-throughput data loading. Since it tracks each Milvus release and publishes its compatibility clearly, teams already running or evaluating Milvus have a supported, obvious entry point.