ReleaseBytes
aws AWS What's New ·

Amazon Bedrock adds request-level usage attribution

aiawsgaengineeraws-bedrockaws-iam
feature

Amazon Bedrock now supports request-level usage attribution for InvokeModel and InvokeModelWithResponseStream APIs, allowing granular tracking of inference usage across teams and applications. This enhancement provides deeper visibility into consumption patterns, aids cost optimization, and simplifies internal reporting without requiring new resources. The feature is available in all Amazon Bedrock commercial regions and builds upon existing attribution capabilities.

  • →Enable request-level usage attribution for InvokeModel APIs
  • →Consistent request-level tagging across Bedrock runtime
  • →Enable model invocation logging to use the new feature
Features (1) ›
  • Enable request-level usage attribution for InvokeModel APIs

    Amazon Bedrock customers can now attribute model inference usage to specific teams, applications, environments, and experiments at the individual request level for InvokeModel and InvokeModelWithResponseStream APIs. This provides fine-grained visibility into usage distribution, helping optimize spend and report consumption patterns.

Enhancements (1) ›
  • Consistent request-level tagging across Bedrock runtime

    This release extends request-level metadata capabilities, previously available for Converse and ConverseStream APIs, to the InvokeModel and InvokeModelWithResponseStream APIs. This ensures a consistent method for tagging inference calls across the entire bedrock-runtime endpoint.

Notes (1) ›
  • Enable model invocation logging to use the new feature

    To utilize request-level attribution, customers must enable model invocation logging in their AWS Region and include metadata in inference requests. The feature is available in all AWS commercial regions where Amazon Bedrock is offered.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/05/amazon-bedrock-request-level-usage-attribution/

Related releases