Couchbase, Inc., the developer data platform for critical applications in our AI world, today announced Couchbase 8.0, delivering end-to-end AI data lifecycle support for enterprises building AI applications and agentic systems. With the introduction of three distinct vector indexing and retrieval capabilities that support a variety of diverse vector workloads, Couchbase 8.0 provides a scalable, high performing, AI-ready data platform for building context-aware, real-time AI applications. Available for self-managed and Capella-based deployments, Couchbase 8.0, supports billion-scale vector search with millisecond latency, and tunable recall accuracy, at a low TCO across on-premises, cloud and edge deployments.

“Scaling AI requires a developer database platform built for speed, throughput and reliability. With support for our Hyperscale Vector Indexing (HVI) and end-to-end RAG workflows, Couchbase stands out from other offerings in the market by providing more flexible and comprehensive vector search options,” said Matt McDonough, SVP of product at Couchbase. “By managing the full AI data lifecycle — which spans sourcing and vectorization through LLM engagement, to validation and drift detection — we help customers create trustworthy agentic systems, while reducing latency, boosting recall accuracy and lowering total cost of ownership.”

Couchbase’s performance advantage was demonstrated in independent billion-scale vector benchmark testing commissioned by the company. The platform’s tunable HVI delivered up to 19,057 queries per second (QPS) with 28-millisecond latency at 66% recall accuracy to highlight its performance. Its challenger, a leader in Gartner’s Magic Quadrant for Cloud Database Management Systems, hit 6 QPS at 57% recall accuracy. And when tuned for accuracy, by increasing its scan properties, HVI delivered over 700 QPS, achieving 93% recall accuracy with sub-second response times. When compared against its challenger the speed test is more than 3,100 times faster and the accuracy test performs 350 times more work. Couchbase’s combination of speed, scale and accuracy enables enterprises to handle massive vector workloads without inflating infrastructure costs nor sacrificing quality.

“Couchbase’s new vector search capabilities transform how we deliver context-aware video discovery for enterprises. We’re already using SQL++ and full-text search to query metadata across hundreds of thousands of employee-generated videos, and added vector search capabilities takes this to the next level,” said Ian Merrington, CTO at Seenit. “Our customers can find relevant content based on meaning and context, not just exact keywords. As a Capella customer, we’re excited for Couchbase 8.0 and the scalability and TCO benefits that make it the ideal solution for our AI-powered video platform.”

“The single greatest accelerator for enterprise AI adoption is the simplification of the underlying data stack,” said Saurabh Jha, SVP and global head of data and analytics at Tech Mahindra. “Couchbase 8.0’s launch is a pivotal milestone, collapsing the divide between operational data and the vector capabilities essential for modern AI. For our teams at Tech Mahindra, this means faster development cycles, lower architectural complexity and a direct path to deploying high-performance RAG and agentic AI solutions for our customers”

Optimized Vector Indexing Options for Speed, Scale and Accuracy

Teams creating AI applications and agentic systems face significant challenges managing AI-generated data. In the Couchbase FY 2026 CIO AI Survey, 28% of CIOs cite difficulties in managing or accessing necessary data as a key factor disrupting AI projects, and only 16% have a vector database that can efficiently store, manage and index high-dimensional vector data.

Couchbase 8.0 delivers the comprehensive solution enterprises need through its novel three-pronged indexing approach to vector embedding, indexing, storage and access. It is designed to support various vector retrieval scenarios from those that require very broad vector-based context, to those that can control or adjust prompt variables on a more granular basis. These applications require high-speed vector indexing, massive vector index capacity and millisecond vector retrieval times, enabling agentic systems that are efficient, cost effective and respond as fast as humans expect. The platform supports:

  • Hyperscale vector index that scales beyond a billion vector index records without compromising responsiveness or performance. Built on the DiskANN nearest-neighbor search algorithm using the Vamana directed graph construction algorithm, this approach provides the flexibility to operate across partitioned disks for distributed processing and scaling, and run in-memory for smaller data sets.
  • Composite vector index that supports pre-filtered queries that help scope the specific vectors it seeks. Composite vector indexes can be stored and partitioned similarly to other global secondary indexes in Couchbase, providing efficient vector retrieval for targeted use cases.
  • Search vector index that enables queries for vectors via the search service, supporting hybrid searches that contain vectors, lexical search and structured query criteria within a single SQL++ request. This capability enables sophisticated search scenarios that combine multiple data types and query patterns.

Developers can now achieve better performance, accuracy, memory usage and cost savings through flexible, use-case oriented choices for vector indexes. This rare optionality allows organizations to deploy context-defining vectors optimized for their specific application requirements, whether they need to optimize for query complexity, vector volume or precision.

Enhanced Security and Compliance Protects Mission-Critical AI Applications

Recent data collected by Couchbase revealed that more than a third of CIOs expressed a lack of confidence around meeting security or compliance demands, which have disrupted AI projects. With Couchbase 8.0, enterprises benefit from enterprise-grade security and compliance features.

New out-of-the-box support for native data at rest encryption (DARE) provides built-in security that automatically encrypts data stored on disk and decrypts it when accessed, protecting sensitive data from unauthorized users. Support for the Key Management Interoperability Protocol (KMIP) provides seamless, cross-platform management of encryption keys to further strengthen data security and simplify operations.

Cross data center replication (XDCR) offers bidirectional active-active conflict awareness detection that now extends to mobile buckets. This allows mobile clients to sync to the closest cluster to improve reliability, scalability and performance. It protects against data center failure while ensuring nonsensitive data is replicated between regions and protected data is stored locally. This makes it easy for globally distributed businesses to adhere to regional data compliance regulations while supporting mobile consumption.

New Intelligent auto-failover capabilities ensure customers maintain continuous operations even during system disruptions, such as reaching disk capacity or disappearance of ephemeral (transient) buckets.

Enhanced Experience Improves Developer Productivity

Couchbase 8.0 improves developer productivity with streamlined experiences and support for AI use cases. The platform enables developers unfamiliar with SQL++ to use natural language to query data, and adds a query workload repository for query statistics over time to help troubleshooting. The search engine adds user-defined synonyms for more relevant results.

For DevOps, Couchbase Becomes More Efficient and Responsive to Failures

The new release removes friction for developers by reducing operational complexity and minimizing troubleshooting and infrastructure management overhead. It addresses both runtime performance and ongoing data management burdens tied to building agentic systems operating at scale.