StarRocks 4.0 Delivers 60% Faster Analytics and Unified Governance to Power Enterprise Intelligence

CelerData, a fast, secure lakehouse engine for customer-facing and agent-driven analytics powered by StarRocks, announced the availability of StarRocks 4.0, the latest major release of the open-source analytical database project that powers high-performance analytics for modern enterprises.Model training, inference, and AI agent development are creating new challenges for data platforms – demanding not only faster performance and higher concurrency, but also real-time access to governed data across their data and AI assets. Following a sneak peek reveal at the first StarRocks Global Summit, StarRocks 4.0 rises to meet these challenges, combining open data lake flexibility with extreme performance and unified governance, setting a new benchmark for speed, consistency, and scalability in the open analytics ecosystem.

Performance remains at the core of StarRocks’ innovation. Version 4.0 delivers up to 60% faster performance Y/Y, driven by deep optimizations across JOIN, aggregation, and spill handling operations. The redesigned execution engine improves parallelization and reduces CPU consumption, allowing complex, multi-table joins and COUNT DISTINCT queries to run faster and more consistently across both internal and external catalogs.

Flat JSON V2 takes performance further by enabling JSON data to be queried as efficiently as structured data. With global dictionaries, zone map indexes, and late materialization, JSON queries now achieve 3-15X faster execution without flattening or pipeline changes. This means teams can run real-time analytics on logs, clickstreams, and user profiles without sacrificing performance.

“StarRocks 4.0 represents a major milestone in our mission to make analytics both open and performant,” said Sida Shen, product manager, CelerData and StarRocks. “With this release, we are delivering a unified analytics engine that is not only faster, but also more predictable, governable, and interoperable with the open data ecosystem. Whether you are building Agents, real-time dashboards, processing time-series data, or operating at petabyte scale, StarRocks 4.0 gives you the control, speed, and reliability to power modern enterprise intelligence.”

StarRocks 4.0 also makes real-time analytics directly on object storage more efficient at scale. Smarter file bundling, metadata caching, and compaction strategies reduce cloud API calls by up to 90%, thereby lowering operational overhead while maintaining freshness and low latency.

Open data architecture with Apache Iceberg and unified governance
With version 4.0, StarRocks brings warehouse-grade discipline to Apache Iceberg, integrating the openness of a data lake with the performance of a data warehouse. Enhancements include full support for hidden partition tables, optimized Iceberg writes that produce larger, query-optimized files, and a new Compaction API that merges small files on demand for sustained query performance.

Optimized metadata parsing, lightweight statistics collection, and smarter refresh mechanisms eliminate redundant scans and keep Iceberg tables query-ready. Even when table statistics are missing or stale, StarRocks generates cost-efficient execution plans, ensuring consistent query speed across large and evolving datasets.

Beyond technical efficiency, governance has been elevated to enterprise scale. StarRocks 4.0 now supports catalog-centric access control that propagates identity and permissions seamlessly across environments. With JWT-based session catalog integration for the Iceberg REST Catalog and temporary credential vending across AWS, GCP, and Azure, user access is unified, and storage credentials no longer need to be hardcoded. The result is simpler compliance, stronger security, and reduced administrative effort across multi-cloud deployments.

Enterprise reliability and precision at scale
As analytics evolve into mission-critical workloads, StarRocks 4.0 introduces new capabilities designed for precision, reliability, and operational resilience. The addition of Decimal256 arithmetic ensures absolute accuracy for large-scale financial and quantitative workloads, making it ideal for currency settlement, trade reconciliation, and risk modeling.

New multi-statement transaction support brings atomicity to multi-table pipelines, simplifying complex workflows and ensuring analytical consistency across datasets. Meanwhile, ASOF JOIN introduces high-performance time-series joins based on timestamps or sequence IDs, enabling applications in finance, IoT telemetry, and AI pipelines.

Operational improvements strengthen StarRocks in production environments. Node blacklisting helps isolate unstable nodes automatically to improve fault tolerance, while case-insensitive identifiers make migrations easier across mixed environments. Global connection IDs add traceability for distributed debugging, giving operators full visibility into query behavior across clusters.
Together, these enhancements extend the reliability and precision of StarRocks into the most demanding data-driven industries, from financial services and payments to manufacturing, IoT, and AI.

Availability
StarRocks 4.0 is available as part of the open-source project, with enterprise support provided by CelerData.