Amazon SageMaker AI now supports OpenAI-compatible APIs for inference endpoints

Amazon Web Services has announced that SageMaker AI now supports OpenAI-compatible APIs for inference endpoints, allowing developers to connect existing tools and frameworks directly to SageMaker without code modifications. The integration enables developers using the OpenAI SDK, LangChain, and Strands Agents to simply change their endpoint URL while maintaining their existing authentication approach, SDK calls, streaming logic, and framework integrations. The feature leverages existing AWS credentials with automatic token refresh, eliminating the need for custom integration code or API format changes. The new capability allows organizations to maintain their current development workflows while gaining access to SageMaker's infrastructure benefits, including custom GPU instance selection, private VPC deployment, support for open source and fine-tuned models, and configurable auto-scaling policies. The OpenAI-compatible API support is immediately available across 14 AWS regions, including major markets in North America, Europe, Asia Pacific, and South America.