Amazon Bedrock integrated OpenAI's GPT-5.6 models (Sol, Terra, and Luna) into its bedrock-runtime endpoint, supporting Responses, Converse, and Chat Completions APIs. This allows unified management of OpenAI API requests under Bedrock's existing controls, including logging and cost tracking via AWS tools. Bedrock further enabled cross-Region inference, automatically routing requests within or across AWS Regions for improved throughput and reduced inference costs without manual capacity management.
Cross-Region inference supports geography-based routing to maintain data locality and global routing for broad capacity access, with Global cross-Region inferencing priced lower per token than other methods. This feature became available in all AWS Regions offering OpenAI models on Bedrock. Users should consult Amazon Bedrock documentation for model specifics and regional availability.
Source & further reading
Amazon Bedrock expands API support and introduces Cross Region Inferencing for OpenAI models ↗Source published 2026-08-17. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.