Simplify model selection in Amazon Bedrock with the open source Model Profiler

This article introduces the Amazon Bedrock Model Profiler, an open source tool that consolidates model discovery, comparison, and selection across 120+ foundation models available on Amazon Bedrock.

• Aggregates model metadata from seven AWS APIs and external sources into a single searchable interface
• Provides model explorer with filtering by provider, capabilities, modalities, and regional availability
• Enables side-by-side comparison of up to 25 models across pricing, specifications, and regional availability
• Displays regional availability matrix across 33 Amazon Bedrock-enabled regions with lifecycle status
• Runs fully automated serverless pipeline using AWS Step Functions and Lambda with 97% cache hit rate
• Offers local deployment (under 5 minutes, no cloud infrastructure) or AWS serverless deployment with daily refresh
• Supports filtering by consumption options: In Region, Cross-Region Inference Service, Batch, and Mantle

The Model Profiler reduces model evaluation time from hours to minutes, helping teams make data-driven decisions on foundation model selection for production AI applications.