Zilliz Lowers Cost of Billion-Scale Vector Search with Milvus 2.6.x Cloud Update

Event summary

  • Zilliz released Milvus 2.6.x on Zilliz Cloud on January 20, 2026, offering cloud-native optimizations for vector search.
  • New three-layer tiered storage reduces costs by up to 87% with 90%+ cache hit rates.
  • Index Build Level feature allows automatic balancing of recall, performance, and storage efficiency.
  • Hybrid BM25-optimized full-text and vector search delivers up to 7× faster performance than Elasticsearch on selected datasets.
  • Expanded data types and multilingual retrieval support enable more complex global AI applications.

The big picture

Zilliz's update addresses critical bottlenecks in AI infrastructure as organizations scale from prototypes to production. The focus on cost reduction and performance optimization positions Zilliz Cloud as a key player in the rapidly evolving AI infrastructure market. With over 10,000 organizations already using Milvus, the company is well-positioned to capture a significant share of the growing demand for vector search solutions.

What we're watching

Adoption Pace
How quickly enterprises will migrate from prototypes to production using Zilliz Cloud's cost-efficient infrastructure.
Competitive Response
Whether competitors like Elasticsearch will accelerate their vector search capabilities to match Zilliz's performance claims.
Market Expansion
The pace at which Zilliz can expand its customer base beyond the current 10,000 organizations worldwide.

Related topics

Product Launch