SAN FRANCISCO, Dec. 17, 2024 (GLOBE NEWSWIRE) — Zilliz, the company behind the open-source vector database Milvus, today announced the release of Milvus 2.5, which includes powerful advancements to hybrid vector and keyword search capabilities that process queries 30 times faster than traditional solutions while eliminating the need for separate vector and keyword search systems. Available now, Milvus 2.5 marks a significant advancement in making sophisticated AI-powered search accessible to organizations of all sizes.
Traditionally, organizations implementing hybrid search have been forced to maintain two separate systems – one for semantic search and another for keyword search – resulting in duplicate infrastructure, complex integrations, and increased operational costs. Milvus 2.5 solves this challenge through an innovative approach that unifies both capabilities in a single, high-performance platform.
“We’re witnessing a fundamental shift in how organizations approach search,” said Charles Xie, founder and CEO of Zilliz. “With Milvus 2.5, we’re not just combining two search approaches – we’re revolutionizing enterprise search with a solution that’s 30 times faster while dramatically simplifying infrastructure. This is a game-changer for organizations building AI-powered applications.”
Unmatched Speed and Performance
Milvus 2.5 leverages Sparse-BM25 technology, a sparse vector implementation of the BM25 algorithm used by Elasticsearch and other full-text keyword search systems. This approach to hybrid vector-keyword search has produced overwhelming results; with 1 million vectors, Elasticsearch takes 200 milliseconds when tested on fully-managed Elastic Cloud, while Milvus 2.5 takes just 6 milliseconds to return search results on fully-managed Zilliz Cloud. That translates to a performance improvement greater than 30X. Unlike basic hybrid implementations, Milvus 2.5 maintains high accuracy even as document collections grow and change over time, automatically adapting to new terminology and specialized vocabulary across any industry or domain.
Milvus 2.5’s hybrid search provides benefits including:
- Unified Infrastructure: Managing one system rather than two minimizes operational complexity while saving organizations time and resources, including less context switching and no need to master two different sets of APIs.
- Intelligent Query Processing: One request can execute both semantic and full-text search tasks, eliminating two different API calls to separate systems.
- Consolidated Data Management: A unified table structure stores dense (vector-based) and sparse (keyword-based) data alongside shared metadata labels rather than utilizing two separate systems and storing metadata labels twice.
- Enhanced Security and Access Control: With only one system to manage, all access controls are centrally administered, strengthening security compliance and consistency.
For more information on Milvus 2.5 and its groundbreaking features, visit the Milvus website and read our blog post.
About Zilliz
Zilliz is a leading vector database company, founded by the engineers who created Milvus, the world’s most widely-adopted open source vector database. Zilliz’s next-generation database technologies help organizations rapidly create AI/ML applications and unlock the potential of unstructured data. By simplifying complex data infrastructure management, Zilliz is committed to bringing the power of AI to every corporation, organization, and individual.
Headquartered in Redwood Shores, CA, Zilliz is backed by prestigious investors, including Aramco’s Prosperity7 Ventures, Temasek’s Pavilion Capital, Hillhouse Capital, 5Y Capital, Yunqi Partners, Trustbridge Partners, and others. Zilliz’s technologies and products help over 10,000 organizations worldwide easily create AI applications in various use cases. Learn more at zilliz.com or follow @zilliz_universe.
Media Contact
Chris Churillo, VP of Marketing
[email protected]
A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/238441fc-16af-4c35-a90e-2e292b1b7e17
