AWS 博客：按每个正确答案成本选择 Bedrock 上的 OpenAI 模型
原标题：Beyond the price per token: Choosing the right OpenAI model on Amazon Bedrock for your workload
AI 摘要
AWS Machine Learning Blog 发布一篇技术文章，介绍一个开源基准测试工具 openai-on-aws/benchmarks-openai，用于在 Amazon Bedrock 上对比 OpenAI 模型（gpt-5.6-luna、gpt-5.6-terra、gpt-5.6-sol）与 OpenAI API 上两个高性价比模型（gpt-5.4-mini、gpt-5.4-nano）。文章主张按“每个正确答案的成本”而非“每 token 价格”选型，并指出在 agent 场景中轮次增加会导致累计输入 token 近似二次增长。结果显示，在关闭推理的配置下，luna 在 AIME 上每个正确答案成本为 0.0021 美元，低于 mini 的 0.0139 美元。
正文节选
Beyond the price per token: Choosing the right OpenAI model on Amazon Bedrock for your workload Organizations building generative AI applications usually compare models the same way: dollars per million tokens. It’s the number on every pricing page, so it becomes the number in every spreadsheet. But production workloads don’t buy tokens. They buy outcomes: a resolved support ticket, a completed research brief, a correct financial summary. Between the pricing page and the outcome sit multipliers