gpt-6IMAGE
Leaderboard shows Opus 5.5 at 94.2% and Astra at 98%
A leaderboard screenshot shows Claude Opus 5.5 falling below its starting value and being called nerfed in advance, alongside scores for Astra, Sonnet 5.5 and Sol.
Tony Quackprano (❖,❖)
@xvremyaperemenxArchived 2026.10.03Originally posted 2026.10.03
Tony Quackprano (❖,❖) @xvremyaperemenxOriginally posted 2026.10.03
Claude Opus 5.5 showing the significance of the fall.... will Fable 5.5 be released soon? ust yesterday it was higher than its starting figures, and today it shows 94.2% GPT-6 Astra - 98.0% Sonnet 5.5 - 100.9% GPT-6.1 Sol - 106.7% but 94.2% is still invested in the initial results, so they say that the model has already been “nerfed” in advance
x.com/xvremyaperemenx/status/2106088925949219125Archived 2026.10.03 10:04