Amazon Bedrock Model Distillation: Boost function calling accuracy while reducing cost and latency

Amazon has announced the general availability of Amazon Bedrock Model Distillation, a technique that enables organizations to create smaller, more efficient AI models with performance comparable to larger foundation models.

• Transfers knowledge from large teacher models to smaller student models
• Focuses on improving agent function calling accuracy
• Reduces model latency by up to 72%
• Provides significant cost savings with smaller models
• Supports expanded model options like Llama 3.x and Anthropic Claude models

Key benefits include faster inference, lower operational costs, and maintaining high accuracy across various AI agent applications. The technology helps organizations deploy sophisticated AI experiences more economically and efficiently.