gpt-6POST
Opus 5.5 said to lead Astra by a generation on visual creative tasks
The author used nearly a quarter of the weekly Claude Code quota the night they topped up, and says Opus 5.5 is clearly stronger than GPT-6 Astra on visual and creative tasks.
花叔
@AlchainHustArchived 23 days agoOriginally posted 2026.10.10
花叔 @AlchainHustOriginally posted 2026.10.10
完蛋，今晚19点刚刚充值的Claude Code，已经跑了接近25%周额度，还有一大堆终端任务在跑着。 最近在做比较考验视觉和创意类的任务时，确实感觉Opus5.5已经比其他模型又拉出了一代的差距，包括GPT-6 Astra也差很多
x.com/alchainhust/status/2108590803207762147Archived 2026.10.10 10:03