Amazon Bedrock has introduced application inference profiles, allowing organisations to tag on-demand models and monitor costs. This feature enables custom tagging for foundation models, improving cost tracking across AI applications. It integrates with AWS services for better cost visualisation and management. The profiles can be created using model ARNs or by copying existing ones, and can be dynamically retrieved using Resource Groups API. This feature is now available in all AWS Regions where Amazon Bedrock is offered, enhancing cost allocation for generative AI initiatives.

**For organizations using AWS services:** You can now precisely track and allocate costs for AI model usage across different projects or departments. This granular visibility allows you to optimize spending, justify investments, and make data-driven decisions about AI resource allocation. By tagging models with custom identifiers, you'll gain insights into which AI applications are driving value and which may need reassessment. The integration with existing AWS cost management tools means you won't need to overhaul your financial processes to accommodate this new capability.

As AI becomes more pervasive in business operations, expect cost management features like this to become standard across cloud providers. You'll likely see the emergence of AI budget specialists who can optimize spending across multiple platforms and model types. Prepare for increased pressure to demonstrate ROI on AI investments, as these tools make it easier to scrutinize costs. In the near term, you may need to update your financial reporting processes to incorporate this new level of detail on AI expenditures. Looking ahead, this granular cost data could fuel the development of AI performance metrics that go beyond pure financial measures, potentially linking model usage to business outcomes.