
Amazon OpenSearch Serverless now supports Binary Vector and FP16 compression, reducing memory requirements and lowering latency while maintaining acceptable accuracy for search and analytics workloads.


<div>
<p>
Amazon has expanded OpenSearch Serverless with new features that help reduce costs and improve performance:
</p>
<ul>
<li>Added Binary Vector and FP16 compression techniques</li>
<li>Reduces memory requirements and lowers costs</li>
<li>Decreases latency while maintaining acceptable accuracy</li>
<li>Measures compute capacity in OpenSearch Compute Units (OCUs)</li>
<li>Now available in 17 global regions, including US, Asia Pacific, Europe, and AWS GovCloud</li>
</ul>
<p>
These enhancements make OpenSearch Serverless more cost-effective and efficient for search and analytics workloads, simplifying infrastructure management for users.
</p>
</div>


Amazon OpenSearch Serverless now supports Binary Vector and FP16 cost savings features

Related articles

Related articles

Oct 8
2024
Amazon OpenSearch Serverless introduces a suite of new features and enhancements

Sep 18
2025
Amazon OpenSearch Serverless now supports Disk-Optimized Vectors

Aug 12
2025
Amazon OpenSearch Serverless now supports kNN Byte vector and new data types

Aug 7
2025
Amazon OpenSearch Serverless adds support for Hybrid Search, AI connectors, and automations