
- Who
- AWS
- What
- made generally available cross-Region inference routing in Amazon SageMaker AI
- When
- Quote
The cross-Region inference routing in Amazon SageMaker AI is generally available starting November 18, 2026, in all commercial Regions where inference is supported for real-time endpoints.
What happened
Amazon SageMaker AI made cross-Region inference routing generally available on November 18, 2026, for real-time endpoints in supported commercial Regions. Salesforce integrated the router into its Internal Gateway to direct eligible requests to endpoints in other Regions and accounts while keeping residency-bound requests local. In a failover test, about 78 percent of probe traffic shifted to a remote Region after local requests began returning errors; Salesforce plans further testing before broader adoption.
Why it matters
Salesforce can route eligible requests to capacity in other Regions while keeping residency-bound traffic local; the test shifted about 78 percent of probe traffic after local requests began returning errors.
The story so far
AWS launched cross-Region inference routing for Amazon SageMaker AI on November 18, 2026, making the feature generally available for real-time endpoints across all supported commercial Regions. This routing capability enables users to automatically direct eligible inference requests to endpoints in different Regions or accounts, while ensuring that requests restricted by data residency requirements remain local. Salesforce has integrated the router into its Internal Gateway to manage this traffic. In a recent failover test, Salesforce found that approximately 78 percent of its probe traffic shifted to a remote Region after local requests began returning errors. The company stated it plans to conduct further testing before adopting the routing feature more broadly.
