Summary
On July 13, 2026, AWS made OpenAI's GPT-5.6 model family -- Sol (flagship), Terra (balanced), and Luna (cost-efficient) -- generally available on Amazon Bedrock, priced at OpenAI first-party rates with usage counting toward AWS commitments. The models support prompt caching with explicit cache breakpoints for a 90% discount on repeated context and are callable via the Bedrock Console or the Responses API on the bedrock-mantle endpoint.
What changed
AWS made OpenAI GPT-5.6 Sol, Terra, and Luna generally available on Amazon Bedrock at OpenAI first-party pricing that counts toward AWS commitments, with prompt caching via explicit cache breakpoints (90% discount) and access through the Bedrock Console or the Responses API on the bedrock-mantle endpoint; Sol in US East (N. Virginia/Ohio), with Terra and Luna also in US West (Oregon).
Why it matters
Landing OpenAI's newest frontier family on Bedrock at first-party rates with committed-spend eligibility lets AWS-anchored enterprises adopt GPT-5.6 without leaving their governance, billing, and networking boundary, reinforcing the multi-cloud model-marketplace pattern where the same model ships across competing clouds within days of launch.
Evidence excerpt
GPT-5.6 supports prompt caching with explicit cache breakpoints, so repeated context across agentic workflows is billed at a 90% discount ... accessed using the Amazon Bedrock Console or the Responses API on the bedrock-mantle endpoint.