Vector database for enterprise-grade artificial intelligence applications company Zilliz Inc. today announced the release of Milvus 2.6, a new version of its open-source Milvus vector database that significantly reduces costs associated with artificial intelligence applications.

The new release includes tiered hot/cold storage and advanced vector compression techniques that allow users to scale up efficiently without compromising performance. For organizations, the release helps keep infrastructure and storage expenses under control.

New features in Milvus 2.6 include “Write-Ahead Log with Woodpecker,” a service that eliminates dependency on external message queues like Kafka. The feature is built with diskless architecture and has been designed to simplify operations, lower infrastructure costs and enhance data ingestion speed.

Milvus 2.6 also simplifies operational management with new native APT/YUM package deployments that allow users to install and maintain databases more easily. Complementing the package deployment functionality are new built-in streaming nodes that streamline real-time data ingestion to make external integrations redundant. The idea here is to allow organizations to focus more resources on innovation rather than managing database operations.

Developers are not forgotten, with the new release including new built-in tools that support functions such as the direct ingestion of raw content, including text, images and audio. Doing so removes the need for external preprocessing pipelines and hence, speeds up development cycles. Additionally, advanced text and JSON search capabilities, including optimized indexing and queries, further simplify the development of sophisticated applications.

“For over a year, I’ve emphasized that cost reduction is critical for widespread AI adoption,” said Zilliz Chief Executive Charles Xie. “As data volumes continue to explode, organizations need vector database solutions that can scale efficiently while keeping costs under control. Milvus 2.6 furthers our commitment to this vision with innovations that streamline infrastructure requirements and optimize resource utilization.”

Zilliz notes that transparency and flexibility remain its core values, as organizations can freely customize, audit and contribute back to the project. Milvus is now used by over 10,000 organizations globally, powering large-scale AI applications across diverse industries.

Image: SiliconANGLE/Reve