GPT-6 Sol 与 Luna 公布在 FrontierCode、DeepSWE 和 AutomationBench 上的成绩与单任务成本，Sol 与 Fable 5.1 得分接近但每任务便宜约 80% 至 85%。
原文并列给出 GPT-6 Sol 与 Luna 在三个基准上的得分和单任务成本，可以顺着价格与成绩的取舍做比较。
Benchmarks GPT-6 Sol & Luna:
• FrontierCode: 48.4% vs. Fable 5.1’s 48.7%, both at xhigh. $1.37 vs. $9.27 per task, roughly 85% cheaper.
• DeepSWE: 68.8% at max vs. Fable 5’s best result of 69.9% at xhigh. $2.74 vs. $13.41 per task, roughly 80% cheaper.
• AutomationBench: 33.2% at xhigh vs. Fable 5.1 with Opus 5 fallback at 31.4% on max. $0.27 vs. at least $2.45 per task.
GPT-6 Luna:
• DeepSWE: 66.6% at max vs. Fable 5’s 65.4% at medium. $0.22 vs. $6.09 per task, roughly 96% cheaper.
API pricing per million input/output tokens:
• Sol: $2 / $10
• Luna: $0.10 / $0.50
Pricing: "GPT-6 Sol will cost $2 per million input tokens and $10 per million output tokens, down 50% from GPT-5.6 Sol's promotional prices, while GPT-6 Luna will cost $0.10 per million input tokens and $0.50 per million output tokens." Astra realiability at much much cheaper price在 X 查看被引用的帖子
来源：@kimmonismus · x.com