What you'll learn

Getting consistent, cost-effective results from Amazon Bedrock requires more than just sending a prompt to a model. It demands deliberate choices about which model to use, how to configure its behavior, and how to structure what you send it. In this course, Amazon Bedrock Parameters and Model Behavior, you'll learn to control every lever that shapes a foundation model's output inside Amazon Bedrock. First, you'll explore the foundation model landscape available in Bedrock, including Amazon Nova, Anthropic Claude, Meta Llama, and Mistral families. Next, you'll discover how inference parameters like temperature, top-p, and max tokens work under the hood, and you'll see firsthand how adjusting them changes response randomness, length, and determinism. Finally, you'll learn how prompt and parameter choices directly affect your Bedrock bill, understanding token based pricing and making targeted adjustments to control spend without degrading outcomes. When you're finished with this course, you'll have the skills and mental models to configure Amazon Bedrock for production workloads, selecting the right model, dialing in parameters, crafting reliable prompts, and keeping costs predictable.