Announcing General Availability of Real-Time Mode for Spark Declarative Pipelines on Databricks
Databricks, Thursday, October 29th, 2026
Databricks announces general availability of Real-Time Mode for Spark Declarative Pipelines, with millisecond end-to-end streaming latency.
Real-Time Mode (RTM) is an evolution of Spark Structured Streaming that streams continuously and runs all query stages at once, reaching millisecond latencies for operational workloads such as fraud detection, live personalization and real-time features.
It is now supported in Spark Declarative Pipelines, which handle dependency resolution, incremental processing, orchestration, recovery and data quality.
ForEach sinks allow instant writes to operational stores.
Databricks says this removes the need for a separate engine like Apache Flink and reports RTM up to 92% faster than Flink on customer feature computation workloads.