gpt-6IMAGE
多模型代码审计评审：Astra 领先，GPT-6 Sol 垫底
让多个新模型做代码审计再由评审团打分，Astra 最佳，Opus 与 Fable 紧随，Grok 4.7 和 GPT-6 Sol 明显落后。
Theo - t3.gg
@theo2026.09.25收录原推发布于 2026.09.25
Theo - t3.gg @theo原推发布于 2026.09.25
Had all the fancy new models do a big audit for me, then set up a judge panel to rank the quality of their audits. Astra slaughtered. Opus and Fable close behind. Grok 4.7 and GPT-6 Sol were WAY behind.
x.com/theo/status/2103242875320615308采集于 2026.09.25 10:01