Astra 6.2 浏览器操作表现突出,智能体任务却被削弱
OpenAI 新版 Astra 6.2 有点偏科:浏览器操作很强、token 也省,但智能体任务被削弱,Sol 甚至不如 Sol 6.1。如果你在选智能体模型,先别急着升级,看看旧版本是否更稳。
OpenAI 的 Astra 6.2 在浏览器操作类任务上表现突出,token 用量也极为高效,但智能体任务的表现被明显削弱,被认为是 OpenAI 为压低成本牺牲了部分性能,实际效果并不理想。另一个版本 Sol 则更令人失望,成绩落后于前代 Sol 6.1,出现升级反而退步的情况。
从评测结果看,Astra 6.2 的定位更像一个专精浏览器场景的模型:在浏览器操作上它能拿到领先的分数,且极高的 token 效率意味着同样的任务成本更低。但智能体任务所需的规划与多步执行能力被削减,使其在复杂自动化工作流中不如预期。Sol 6.1 优于新发布的 Sol,说明同批次更新中并非所有版本都值得迁移。
对开发者而言,这次更新的启示是升级不等于全面受益:如果主要场景是浏览器自动化,Astra 6.2 可能更省钱更好用;如果依赖智能体编排和长链条任务,留在旧版本或选择其他竞品可能更稳妥。
🚨 Astra 6.2 Beats On Browser Use But Disappoints on Agentic Tasks
OpenAI appears to be sacrificing performance for cost, and it doesn't quite work.
- nerfed on agentic tasks - excels at browser use - extremely efficient on token usage
Sol is even more disappointing and comes behind Sol 6.1.
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 AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Astra 6.2, OpenAI, agents, browser automation.
Source attribution: curated from Bindu Reddy (@bindureddy).