gpt-6IMAGE
Claude Opus 5.5 发布，Terminal-Bench 超过 GPT-6 Astra
图表显示 Opus 5.5 全面领先 Fable 5.1，在 Terminal-Bench 上以 66.4% 对 57.9% 胜过 GPT-6 Astra，科研能力近乎翻倍。
Dan
@DanDr1s2026.09.23收录原推发布于 2026.09.23
Dan @DanDr1s原推发布于 2026.09.23
🚨 Claude Opus 5.5 is OUT. Beats Fable 5.1 on EVERY benchmark in this chart. Even beats GPT-6 Astra on Terminal-Bench: 66.4% vs 57.9%. And scientific research nearly doubles over Opus 5: 29% → 58.7%. Okay Anthropic 👀
x.com/dandr1s/status/2102434916239917224采集于 2026.09.23 10:28