Announcing OpenAI-compatible API support for Amazon SageMaker AI endpoints
TL;DR
Amazon SageMaker AI adds OpenAI-compatible API support for real-time inference endpoints. Users of the OpenAI SDK, LangChain, or Strands Agents can now invoke models on SageMaker AI by changing only the endpoint URL — no custom client, SigV4 wrapper, or code rewrites required. The launch makes existing OpenAI-style code work directly against SageMaker endpoints, lowering migration cost between the two platforms.
Nauti's Take
For teams already on the OpenAI SDK this is a clear win: SageMaker endpoints can be wired up without refactors, making multi-cloud and failover setups cheaper to build. The catch: OpenAI-compatible doesn't mean feature-identical — tool calling, streaming edge cases, and model quirks still need per-endpoint testing.
Most useful for AWS-centric teams wanting OpenAI ergonomics without API lock-in; less relevant for shops already deep in Bedrock or a single cloud.
Sources
