7 Years, 2 Major Rebuilds, 40K+ GitHub Stars: The Rise of Milvus as the Leading Open-Source Vector Database

What this post added

This post chronicles the 7-year journey of Milvus, highlighting two major rebuilds and its growth to over 40,000 GitHub stars. It details key milestones including the open-sourcing of Milvus 0.10, joining the LF AI & Data Foundation, winning the BigANN challenge, and the ground-up reinvention for Milvus 2.0. It then focuses on the significant advancements in 2025 with Milvus 2.5 introducing native hybrid search and Milvus 2.6 delivering substantial performance and cost improvements through features like tiered storage, 1-bit quantization (RaBitQ), an improved full-text engine, Path Index for JSON, AiSAQ, Geospatial Search with R-Tree, CAGRA+ Vamana, a streamlined data workflow, and increased collection support. The post also introduces two new open-source tools: DeepSearcher for local deep research and Claude Context for enhanced AI coding assistants. Finally, it emphasizes Milvus's adoption by over 10,000 teams in production for various AI workloads, citing performance metrics like sub-50ms retrieval and multi-tenant capabilities.