The next generation of Amazon OpenSearch Serverless is now generally available

AWS has announced the general availability of the next generation Amazon OpenSearch Serverless, a fully managed search and vector engine specifically designed for customers building AI agents. The updated platform delivers significant performance improvements, scaling 20 times faster than its predecessor and provisioning resources in seconds to handle unpredictable agentic workflows. The service now features scale-to-zero capabilities with pay-per-usage pricing, allowing customers to save up to 60% compared to provisioning OpenSearch clusters for peak loads. The architectural improvements center around complete decoupling of compute and storage through a new shared storage layer, enabling independent scaling of compute resources while maintaining instant readiness for traffic spikes. AWS has simplified network connectivity by introducing two resource-based endpoints—collection level and regional—that streamline multi-VPC and on-premise connectivity using standard VPC APIs. The platform also launches with native integrations for AI development platforms including Vercel and Kiro, allowing developers to provision search infrastructure directly from their development environments using natural language commands, and integration with OpenSearch Agent Skills for popular coding platforms like Claude Code, Cursor, and Codex.