OpenAI 发布 GPT-6.2 Astra,成本降至 GPT-6 Astra 三分之一以下
OpenAI 把 GPT-6.2 Astra 定位成干重活的新旗舰,官方说评分新高还便宜到前代三分之一不到。做幻灯片、网页、3D 设计这类任务的开发者可以先测测实际效果和账单,再决定要不要从 GPT-6 Astra 迁移。
OpenAI 发布新模型 GPT-6.2 Astra,定位为面向高难度项目的旗舰模型。官方称该模型在专业工作和科学研究类评估中创下新高,同时每任务成本不到 GPT-6 Astra 的三分之一。
从产品描述看,GPT-6.2 Astra 的能力提升主要体现在幻灯片、网页和 3D 设计等产出类任务上,也就是模型直接交付可用成品的场景,而不只是回答问题。成本下降叠加评估成绩上升,意味着同等预算下可以跑更多任务轮次。
对开发者和企业用户来说,关键变量是性价比:如果成本真的降到前代的三分之一以下,复杂任务的批量调用成本结构会被改写。具体基准名称和对比数据官方推文中未给出,需等完整评测报告。
GPT-6.2 Astra发布 看评分是遥遥领先🙂 OpenAI @OpenAI GPT-6.2 Astra—a frontier model built for your hardest projects. It sets new highs on evals across professional work and scientific research at less than a third of the cost per task of GPT‑6 Astra. You’ll see the difference in things like slideshows, websites, and 3D designs. 🔗 View Quoted Tweet📊 2 ⚡
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include GPT-6.2 Astra, OpenAI, reasoning models, coding assistants.
Source attribution: curated from 小互 (@imxiaohu). Cross-source corroboration: 11 independent sources have covered this development (including @koltregaskes, OpenAI, Vaibhav Srivastav), confirming industry-wide signal strength.