Amazon SageMaker AI Async Inference now supports inline request payloads

Short summary

Amazon SageMaker Async Inference now supports inline request payloads, enabling direct data submission in InvokeEndpointAsync API calls instead of requiring S3 uploads. This eliminates an extra data staging step from inference pipelines and reduces operational overhead. Particularly valuable for workloads with frequent inference calls, strict latency requirements, or applications seeking simpler deployment and lower data transfer costs.

• •Inline payload support in SageMaker Async Inference API
• •Eliminates S3 upload requirement before each invocation
• •Simplifies inference pipelines and reduces operational overhead

Generated with AI, which can make mistakes.