Opus 5 High scores 98.81% vs GPT 5.6 Sol Max's 21.42% on ARC-AGI-3 puzzle
On one ARC-AGI-3 puzzle, Claude Opus 5 High scored 98.81% while GPT 5.6 Sol Max scored 21.42%. A Reddit user asked Sol Max to analyze both model outputs and shared ARC-AGI replay links for each attempt.
1 source
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- Claude Code 2.1.227 fixes subscription-tier, Bash and TUI bugs
- Curated resources for the open Agent2Agent protocol
- Suno to cap song downloads to curb AI slop
- Claude Code plugin translates 'Claudish' output into plain English
- Claude Code v2.1.227 fixes flag evaluation and Bash command failures