The Point: OpenAI GPT-5.6 Sol, Terra, and Luna are available on Amazon Bedrock, covering requirements from complex reasoning to cost-effective high-volume inference.
OpenAI is deploying three new model variants of the GPT-5.6 family on Amazon Bedrock. The models are optimized for different requirements: Sol for complex reasoning tasks, Terra for productive applications, and Luna for cost-sensitive workloads.
OpenAI has made GPT-5.6 Sol, Terra, and Luna generally available on Amazon Bedrock. The models build on Amazon’s next generation of inference engine, which is designed for high performance, security, and reliability. Pricing follows OpenAI rates, and usage counts toward existing AWS commitments.
GPT-5.6 Sol is the flagship reasoning model and, according to OpenAI, the most capable to date. Sol achieves 80 points on the Artificial Analysis Coding Agent Index (2.8 points above the next best model), while using less than half the output tokens, requiring less than half the time, and costing approximately one-third less. In the cybersecurity evaluation ExploitBench, Sol scores 73.5% compared to 47.9% for GPT-5.5 at comparable token budget. On the Agents’ Last Exam (evaluation of long-term professional workflows across 55 occupational fields), Sol achieves 53.6 points and outperforms the next best model by 13.1 points. Sol additionally supports “max reasoning effort,” which allows computational power to be increased for complex tasks.
GPT-5.6 Terra is the balanced model for productive everyday use, delivering better performance than GPT-5.5 at lower costs. GPT-5.6 Luna is the fast and cost-effective variant for high-volume inference tasks such as classification, summarization, and routing. All three variants enable model capacity and costs to be tailored to individual workloads.
Amazon’s inference engine manages volatile agent traffic through capacity pooling, which absorbs demand spikes while isolating throughput per customer. In-region inference ensures data sovereignty and meets strict data residency requirements. GPT-5.6 on Amazon Bedrock introduces prompt caching with explicit cache breakpoints to efficiently utilize repeated contexts (system instructions, tool definitions, reference files) across invocations.
Source: aws.amazon.com · Published July 13, 2026
Lumi AI News — AI-assisted curation in accordance with Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.7.3.