Industry story
Amazon Bedrock adds Cross-Region Inferencing for OpenAI GPT-5.6 models
cloud-costs inference model-pricing orchestration
Amazon Bedrock now supports OpenAI's GPT-5.6 model family (Sol, Terra, and Luna) on its bedrock-runtime endpoint, with compatibility across three APIs: Responses, Converse, and Chat Completions. The launch also introduces Cross-Region inference for these OpenAI models, automatically routing requests across AWS Regions for higher throughput and lower costs. Global cross-Region inference is priced lower per token than in-Region access.
Full analysis
Amazon Bedrock now supports OpenAI's GPT-5.6 model family (Sol, Terra, and Luna) on its bedrock-runtime endpoint, with compatibility across three APIs: Responses, Converse, and Chat Completions. The launch also introduces Cross-Region inference for these OpenAI models, automatically routing requests across AWS Regions for higher throughput and lower costs. Global cross-Region inference is priced lower per token than in-Region access.
The integration brings OpenAI models under the same enterprise governance controls already available for native Bedrock models, including invocation logging to S3 or CloudWatch, usage metrics, and spend visibility in AWS Cost Explorer. A new US Geo (US CRIS) option allows data to remain within US geography while still scaling across multiple US Regions. This tightens Amazon's position as a multi-model inference platform, letting agent builders mix OpenAI and other models while keeping observability and cost attribution unified.
Comments