The AWS Machine Learning Blog outlines how to use OpenCode, an open-source terminal-native AI coding agent, together with open weight models available on Amazon Bedrock. The post presents this pairing as a way to build a coding assistant that is secure, flexible, and billed on a pay-per-use basis.
The guide covers configuring multi-model workflows, selecting an appropriate model for each task, and keeping data within the user’s own AWS account without managing underlying infrastructure.
Why it matters
Combining an open-source agent with open weight models on a managed platform gives developers control over model choice and data location while avoiding infrastructure overhead. The pay-per-use model ties cost to actual usage.
Who should care
Developers and teams building AI coding assistants, and organizations that want to keep data inside their own AWS environment, are the primary audience for this walkthrough.