gpt-6IMAGE
Claude Opus 5.5 launches, tops GPT-6 Astra on Terminal-Bench
A chart shows Opus 5.5 leading Fable 5.1 on every benchmark, beating GPT-6 Astra on Terminal-Bench at 66.4% vs 57.9%, and nearly doubling Opus 5 on scientific research from 29% to 58.7%.
Dan
@DanDr1sArchived 2026.09.23Originally posted 2026.09.23
Dan @DanDr1sOriginally posted 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/2102434916239917224Archived 2026.09.23 10:28