AI 导读
Perplexity 宣布 GPT-6 Sol 已向全体用户开放，在其 Wide-And-Deep-Research（WANDR）评测上以五分之一的成本超过 Opus 5。Sol 将担任 Computer 用户的 Light Effort 编排模型，Astra 继续作为 High effort 的编排模型。Aravind Srinivas 同时祝贺 OpenAI 持续推出 pareto-optimal 模型。
推荐理由
Perplexity 给出自家 WANDR 评测的得分与单任务成本对照，读者可据此比较 GPT-6 Sol 与 Opus 5 的性价比。
正文
GPT-6 Sol is now available to all Perplexity users. On our Wide-And-Deep-Research (WANDR) evals, it outperforms Opus 5 at one-fifth the price. Sol will become the “Light” Effort orchestrator for Computer users, while Astra remains the orchestrator for "High" effort. Congrats to @OpenAI for consistently launching pareto-optimal models!
来源：@AravSrinivas · x.com