Primary source歸藏(guizang.ai)
View sourceEvent ·Official first-hand
GPT 6.2 astra 口碑不佳:协作能力弱,编程下滑,但 3D 建模突出
网友对 OpenAI GPT 6.2 astra 的早期反馈普遍偏负面,认为其表现类似当时的 Opus 4.7 或 4.8:文本表达生硬,难以协作,编程能力出现下降。 社区的一个观察是该模型似乎在 3D 领域过度训练,导致通用能力被挤压,但 3D 生成确实强,建模细节和精细度被认为超过 Claude。 如果这些反馈属实,说明 OpenAI 这次在专项能力与通用协作体验之间做了明显取舍。对日常用模型写代码、做协作的用户来说,可能需要等更多基准测试和使用反馈再决定是否切换。
Current read
GPT 6.2 astra 首批用户反馈不太妙:写作不说人话、编程还退步了,但 3D 建模精度 reportedly 超过 Claude。如果你主要拿模型写代码或协作,先别急着升级,等更多实测;做 3D 内容的可以关注一下。
11 sources41° AI heatLast updated 2026/11/18 01:05:25
Evidence
11 sourcesRelated source 1Artificial Analysis
View sourceRelated source 2TestingCatalog
View sourceRelated source 3宝玉
View sourceRelated source 4Decoder
View sourceRelated source 5The Rundown AI
View sourceRelated source 6Tibor Blaho
View sourceRelated source 7IT之家
View sourceRelated source 8Simon Willison’s Weblog
View sourceRelated source 9OpenAI Blog
View sourceRelated source 10Yuchen Jin
View sourceConflict check
Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.