Report
Claude Opus 5.5 reportedly takes the top spot on RSI-Exam
A post lists Opus 5.5 at 0.5357 across 88 executable research tasks, ahead of GPT-6-astra at 0.5126.
TLDR
A post says Claude Opus 5.5 ranks first on RSI-Exam and leads on its full, public and private splits. It lists Opus 5.5 at 0.5357 across all 88 executable research tasks, compared with 0.5126 for GPT-6-astra. The post also lists Grok 4.7 at 0.3842, up from 0.3671 for Grok 4.6, and says Gemini 4 is on the way.
Combined views
—
1 Source, first seen ago
